ysmenu on the r4ds guide

http://forums.afterdawn.com/thread_view.cfm/731812

Ubuntu LAMP Drupal

http://drupal.org/node/439204

http://mazesolutions.me/web-development-blog/linuxunix/installing-lamp-on-ubuntu-710-linuxapachemysqlphp/100

<?php echo "Hello World"; ?>

$ sudo /etc/init.d/apache2 restart

$ sudo /etc/init.d/apache2 stop

http://ubuntuforums.org/showthread.php?t=550454

To remove apache2 permanently from startup scripts:
sudo update-rc.d -f apache2 remove

To reinstate apache2 in the startup scripts:
sudo update-rc.d apache2 defaults

http://ubuntuforums.org/showthread.php?t=550454
Edit /etc/default/apache2 as root and set NO_BOOT to 1.

NDS Bookmark

http://www.teamcyclops.com/forum/showthread.php?p=9757

Download moonshell v1.71 from here and don't touch it.

Go here: right here.... and download the latest R4 firmware. Extract it with something like WinRAR.

Ignore the _system_ folder and the _DS_MENU.DAT, take the _DS_MSHL.nds and drag it into the moonshl folder that came with it.

Open the moonshl folder and find the .nds file you just dragged into it. Rename it moonshell.nds instead of _DS_MSHL.nds.

Go to the Evolution moonshell you downloaded and extract it. Open up the PLUGINS folder. Find reset.mse.

Drag it to the R4 moonshell's PLUGINS folder. When it asks you if you want to replace the existing reset.mse say yes.

Now Just take the R4's moonshl folder and drag it to the root of your MicroSD.

Believe

信じる
Shinjiru

E-books

http://www.gutenberg.org

http://listverse.com/2008/01/28/top-10-books-of-all-time/

How to type Korean in Linux

IBus Keyboard Input Method
-----------------------------
System > Administration > Language Support

Keyboard Input Methods "ibus"
Install Languages > Korean
IBus Input Method > Korean > Add

Ctrl+Space to enable/disable (or click the keyboard icon on the toolbar and select Korean)

You have two options: remap your English keyboard (a) and put sticky notes or do a lot of scrolling (b).


Option a: Korean 2bul
-----------------------------
The 'm' in Korean is the character 'a' in the English keyboard, etc.
ㅁ m = korean 2bul (a)
ㄴ n = korean 2bul (s)
ㅎ h = korean 2bul (g)
ㅓ eo = korean 2bul (j)
ㅏ a = korean 2bul (k)

So
한 = han = korean 2bul type 'gks'

This method, you can type Chinese too.


Option b: Hangul Romaja
-----------------------------
Hangul Romaja (안년 하세요 = annyeon haseyo):
안 an
년 nyeon

하 ha
세 se
요 yo

Problem is, I got these to work with a lot of scrolling in the pop-up, the ones I want are always the last one the list!

SCIM (can't get it to work ...)

-----------------------------
Have SCIM Manager installed, Install scim-tables-ko and scim-hangul
    $ sudo apt-get install scim scim-tables-ko scim-hangul
Select your chosen input type in IMEngine > SCIM Global Setup and IMEngine > Hangul (keyboard layout 2bul)
    $ scim-setup
Start SCIM (might need to add in System > Preferences > Startup Applications)
    $ scim
Restart Ubuntu


References
-----------------------------
http://translate.google.com/#ko|ko|annyeon%20haseyo
http://ubuntuforums.org/archive/index.php/t-30517.html
http://webcache.googleusercontent.com/search?q=cache:KikFlNNizUIJ:ubuntuforums.org/archive/index.php/t-30517.html+korean+2bul+in+linux&cd=5&hl=en&ct=clnk&client=ubuntu&source=www.google.com
http://ubuntuforums.org/showthread.php?t=30981
http://en.wikipedia.org/wiki/Keyboard_layout#Hangul_.28for_Korean.29
http://langintro.com/kintro/syllable.htm
http://www.mrbass.org/linux/ubuntu/scim/
http://ubuntuforums.org/archive/index.php/t-205098.html  (Nabi package)

PubMed API

PubMed API

The NCBI provides a web service interface to PubMed. For searching this is a two-step process that first performs the search then retrieves the results. Here's an example in PHP5:


http://hublog.hubmed.org/archives/001518.html

return of the jedi

i'll come back 2 years from now ... Apr 28 2012 , if not sooner

100 credits so far, 9 in progress, 39 transferred = 148 credits!

the fork

was at the fork today, time seems to fly by so fast, was waiting for an e-mail you see ... 3:30 hits
no e-mail received, so had to send the other e-mail and so that's that

west won in the end.

Alpacas

http://purldivers.files.wordpress.com/2009/11/alpacas.jpg?w=488&h=515

miniaturized llamas domesticated for fur

http://en.wikipedia.org/wiki/Alpaca

Love Story Korean Movie Trio

A Moment To Remember
My Sassy Girl
The Classic

http://www.youtube.com/watch?v=xXbwzPDL_MU

... been sappy these couple of days ...

Kina Grannis - Valentine

http://www.youtube.com/watch?v=840NbiFF1zM

Nice, cute vid.

사랑해요!

I love you! Sa-rang-hae-yo! 사랑해요!

Owl

http://numenessence.com/owl.html

"How the Owl came to be recognized as the symbol of wisdom is anyone's guess, but it probably had a lot to
do with the Owl's large searching eyes, with the ability to turn its head 270 degrees; and the nocturnal ability to
see well in the darkest of night, while it silently observes its prey before quietly swooping in for the kill! "

Their innate power of observation, and the need to learn and understand what goes on around them
are the true gifts of the Owl personality, as well as the ability to take time out to rest and relax and enjoy the
leisures of life. While these qualities may cause some to think of Owls as nosey, inquisitive, and indolent
people, this is far from true, regardless of how their probing questions and slow reactions may come off at
times. Owls are economical, deep thinking people and resolving problems is their specialty. For this reason,
many Owls are drawn to fields of study requiring in-depth knowledge and analysis, such as science, research,
academia, mechanical, and investigative work. Owls also tend to love the world of nature, and many have
adeptness for the outdoors, a natural green thumb, and especially enjoy the rain and being around water,
accentuated by the need for cleanliness and purification.


"The Black Swan"
http://en.wikipedia.org/wiki/Black_swan_theory
The term black swan was a Latin expression — its oldest reference is in the poet Juvenal expression that "a good person is as rare as a black swan" ("rara avis in terris nigroque simillima cygno", 6.165). [1] It was a common expression in 16th century London as a statement that describes impossibility, deriving from the old world presumption that 'all swans must be white', because all historical records of swans reported that they had white feathers [2]. In that context, a black swan was something that was impossible, or near impossible and could not exist.

http://www.achieve-goal-setting-success.com/life-planning-workbook.html
see DOPE test
dove, owl, peacock or eagle.

http://www.quiztron.com/tests/animal_describes_per_quiz_112727.htm

What animal describes your personality?

Owl

An owl personality is, like the eagle, a bit of a loner. They tend to enjoy their solitude, and usually avoid others by hiding. Owl personalities tend to be quiet, and often have poorer vision than others. However, their hearing makes up for that. They tend to like the darkness, and are attracted to mysterious things and people. In a book, for example, an owl personality would pay more attention to the character who seems to be hiding a secret from everyone. They tend to dwell on things that hurt them in the past, and think of things they could have done to prevent this. They are cold on the outside, yet romantic on the inside. They enjoy creativity and logic.

Definition of Euphimism

Euphemism, is in speech or writing the avoiding of an unpleasant or indelicate word or expression by the use of one which is less direct, and which calls up a less disagreeable image in the mind. Thus for "he died" is substituted "he fell asleep," or "he is gathered to his fathers"; thus the Greeks called the "Furies" the "Eumenides," "the benign goddesses," just as country people used to call elves and fairies "the good folk neighbours."

Plants final review

ABA (seed development) - antagonistic to GA (seed germination, stem elongation, reserve mobilization) and Ethylene (fruit ripening, senescence)

A reduction of GAs re-establishes an ABA/GA ratio appropriate for suppression
of germination and induction of maturation.

These GAs induce a developmental program that leads to vivipary
in the absence of normal amounts of ABA

Auxin (2,4-D, IAA) - high auxin (cell elongation, Mediates the tropistic (bending) response of bending in response to gravity (gravitropism) and light (phototropism)) promotes Et synthesis

Cytokinins (zeatin) - Stimulates cell division., morphogenesis, conversion of etioplasts to chloroplasts, apical dominance, found in roots and shots

plant morphogenesis:
Aux > CK - rooty
CK > Aux - shooty

GA, CK and AUXIN promotes growth

Sencescence mediated by cytokinin (antagonize):ethylene (promote) balance

leaf abscission is mediated by decrease in auxin -> increase ethylene production -> synthesis of hydrolases -> leaf abscission

http://www.plant-hormones.info
http://www.planthormones.info/characteristics.htm

Seed Physiology
-------------------------------------------
Maturation Drying acts as a “switch”
A “switch” is needed to terminate development and promote the transition
to a germination and growth program

Control of Seed Maturation
Two key regulatory factors are abscisic acid (ABA) and restricted water uptake (negative osmotic potential in seed tissues).

ABA plays a role in (inhibits germination and promotes development - reserve synthesis):
• Encouraging development, including reserve synthesis
• Acquisition of ability to withstand water loss: accumulation of
desiccation protectants
• Inhibition of precocious germination
• Induction of dormancy (some species)
• In this sense, ABA ultimately controls seedling vigour - a seedling’s ability
to emerge and become established over a range of environmental conditions

Early evidence for role of ABA
(1) Embryo culture
(2) vivipary – germination of the immature seed on the plant
(3) Species that exhibit preharvest sprouting

insensitivity to ABA, (so-called ‘response mutants’) - defective in a component of the ABA signaling pathways eg. vp1 (maize) and abi3 (arabidopsis) - transcription regulators
- The mutant seeds are green, less dormant, reduced in their storage reserve content
and desiccation-intolerant. Maturation proteins degraded during development –
germination & growth program inappropriately switched on.

vp1 and abi3:
- Activate ABA-responsive genes during seed development
(e.g. storage-protein- and LEA genes).
- Inhibit the expression of germinative- and post-germinative genes
(eg. those involved in reserve mobilization)
- interacts with FUS3 and LEC1



Technological Aspects of Seed Physiology of Relevance to
Agriculture and Forestry
-------------------------------------------
(1) Problems with deep dormancy (subjected to microbes, need dormancy breaking),
non-synchronous germination (some grow fast) and
poor seedling growth (due to poor seed development, need to overseed)
(2) Malting & Cereal Industry: Problems with preharvest sprouting
(3) Preserving valuable genetic backgrounds through cryopreservation
of zygotic and somatic embryos
(4) Genetic modification of nutritional properties of seeds
(5) Seeds can be used as vehicles to host production of therapeutics
(6) Genetic modification to re-direct developmental processes
(7) Controversial ‘terminator technology’

(1) deep dormancy is bad
Seed Treatments to enhance seed germination or seed performance
1. Seed priming - Allows seeds to absorb enough water to initiate metabolic processes,
but insufficient water to complete germination, eg. drum, osmopriming, matripriming (clay)

2. Film coating - colored seeds, Applies a thin layer of polymeric material to the outside of the seed, Fungicides included in the polymer are bound onto the seed

3. Seed pelleting - Allows irregularly shaped seeds to be planted by machines

imbibe - To absorb or take in as if by drinking

(2) pre-harvesting is bad
pre-harvest sprouting - Germination of the physiologically mature grain on the parent plant, problem in malting and cereal industry; due to the premature loss of dormancy

A relative insensitivity to ABA may underlie sprouting susceptibility

Malting Process (barley grains used for production of beer, hard liquor)
- Carefully controlled water content changes – ultimately encourage enzyme (eg. α-amylase for starch hydrolysis) synthesis
& endosperm modification (flavour production) but discourage rootlet growth
(mobilization & utilization of reserves - reduction in malt yield).

(3) Preserving genes via gene banks
Yet, only about 15 spp. feed the world:
5 cereals: rice, wheat, barley, maize & sorghum
2 sugar plants: sugarcane & sugar beet
3 subterranean crops (potato, sweet potato & cassava)
3 legumes (bean, soybean & peanut)
2 tree crops (banana & coconut)
Just 3 spp. (wheat, rice & maize) : 70% of world’s seed crops

Trend in Agriculture: develop a few cultivars of high-yielding crops:
genetic uniformity at the expense of genetic diversity
(potential for global disaster: susceptibility to disease/environmental change)

Tree Seed Centre (Surrey): all tree seeds for BC: maintained at -20oC

Liquid N2 is a new technique: potential infinite longevity of seeds

Modes of Regeneration:
1. embryo culture - zygotic 2N embryo is excised from seed and grown in media
2. Somatic Embryogenesis - embryoids 1. induction - soak in auxin 2. development - globular, heart, torpedo, no auxin 3. drying - soak in aba 4. cryopreservation - add glycerol (cryopreservant)

(4) genetic modification of nutritional properties of seeds
- needs 9 essential amino acids, AILKMFTWV and micronutrients (vit A, zinc, etc)

Seed storage proteins:
albumins
globulins (Legumin 12S, vicilin 7s, deficient in methionine and cysteine)
glutelins (cereals, lack tryp and threonine)
prolamins (cereals)

Need to modify existing proteins of seeds to improve the composition
of essential aas

Basic strategies
1. Engineering the seed’s aa metabolism in order to
increase the free amount of aa
2. Engineering genes encoding endogenous storage
proteins (but adding extra aa's create a longer and unstable protein subject to degradation)
3. Transfer of genes encoding proteins enriched in
deficient aas (e.g. met-rich storage protein is
transferred into a legume) (but allergens) eg. vit A/golden rice

(5) seeds as medicine vehicles

Oleosin Fusion- Oleosin used as ‘transporter protein’... transports therapeutic protein
to oil body of seeds. Later, therapeutic is easily purified

(6) Genetic modification to re-direct developmental processes

A. Engineering Male Sterility for Hybrid Seed Production
barnase, a ribonuclease - produces sterile canola (no anther)
barstar - a ribonuclease inhibitor - produces fertile canola

B. Delaying Seed Pod Splitting (Dehiscence) to Avoid Seed Shatter

20% of seeds can be lost during harvest of canola

(7) Controversial ‘terminator technology’

terminator technology, is the name given to proposed methods for restricting the use of genetically modified plants by causing second generation seeds to be sterile.

to prevent "unauthorized seed-saving" by farmers. -- so farmers will need to buy seeds every year

http://www.globalissues.org/article/194/terminator-technology

Ethylene
-------------------------------------------
Ethylene effects in plants
• Fruit ripening
– A major ethylene effect that contributed to its
discovery
• Stress responses/wounding (Promotes ethylene synthesis)
• Abscission
• Senescence (biological aging)
• Lateral cell expansion
– Evident during triple response
in seedlings
• Root hair formation

High auxin levels promote Et synthesis

Ethylene is synthesized from methionine (CH3-S) group is recycled in the Yang cycle



AdoMet -----------1---------> ACC -----------2----------> Ethylene
• ACC synthase (1) catalyses the rate limiting step for Et
biosynthesis
– Regulated by environmental stress and auxin
– Unstable and present at very low levels in plant cells
• ACC oxidase (2) limits Et synthesis in tissues that make large
amounts of Et

Ripening is blocked in the rin mutant
• Unable to make climacteric Et

Ethylene response to stress:
- Leaf epinasty – Downward curving of leaves, induced by flooding (low O2 concentration, ACC accumulates)
- Aerenchyma: - Formation of air spaces in the cortex of roots due to low O2 concentration in flood

The triple response
– Ethylene reduces elongation growth
& increases lateral growth
• Inhibition and swelling of the
hypocotyl
• Inhibition of root elongation
• Exaggeration of the apical hook

Sencescence mediated by cytokinin:ethylene balance

Ethephon (Ethylene releasing agents) is used to:
• Ripen apples and tomato
– Useful when fruit are picked green and transported to market
• Accelerate abscission
– Useful for thinning fruit crops
• Reduce elongation growth and promote compactness in flowers

Stuff that inhibits Ethylene:
- low O2
- high CO2
- cool temp.
- Silver (Ag2+)
- 1-MCP
- AVG (not approved)

Ethylene action

ETR1:
- has an N-terminal (Et binding domain)
- histidine kinase catalytic site
- C-terminal receiver domain

• Ethylene receptors are negative regulators of ethylene
response
– Receptors are “active” in the unbound state
– Unbound receptor shuts off the ethylene response pathway
– Ethylene binding deactivates the receptors
• Response pathway proceeds
receptors are like locks and ethylenes are keys

• Ethylene receptors (ETR1, ETR2, ERS1, ERS2, EIN4) are functionally redundant
– Disrupting the regulatory domains of one receptor has no effect
on eliciting a ‘constitutive’ ET response

Two mutants:
(a) Ethylene resistant - mutant not responding to ethylene; has one receptor insensitive to ethylene (one lock has a broken key hole, so you can't open it) because of missense mutation in the binding domain, so Et can't bind to it, so by default the receptor shuts off Et response pathway, so you get a tall mutant, against the short wildtypes, easy to stop, only need one insensitive receptor to stop
(b) Constitutive ethylene response - Ethylene response is permanently activated; multiple receptors needed to be disrupted in the regulatory domain (broken lock, so you can open the door without a key) so it doesn't matter if there's ethylene or not, the receptor doesn't work and so it doesn't shut off the Ethylene response, so you get constitutive response so small mutant vs tall wildtypes, hard to start, need many disrupted receptors to start

Ethylene + Cu2+ -> ETR1 -> inactivate CTR1 -> activates EIN2 -> induce EIN3 -> induce ERF1 -> ethylene response

Plant Transformation – Methods for Introducing
Novel Genes Into Plants
-------------------------------------------
I. Agrobacterium-Mediated Transformation

Causes tumorous outgrowths
on plants: “Crown gall tumors”

Host specificity: Dicots and
gymnosperms, limited number of
monocots

tumour - disorganized growth and continuous cell division

Ti (tumour-inducing) plasmid (~200kb)

Transfer of part of Ti plasmid (the T-DNA; T=transferred) from Agrobacterium to plant

Agro. is only organism capable of inter-kingdom DNA transport! (plants, humans, fungi)



Agro is a “natural plant genetic engineer”

Ti plasmid is a natural plant transformation vector

T-DNA - contains genes that encode for auxin & CK (controls cell division), and opines (modified amino acid) synthesis



Regions of Ti-plasmid:

1) T-DNA: Part of the plasmid that gets physically transferred to
plant (via formation of T-DNA intermediate) - contains Oncogenecity genes: opine, CK & auxin, and left and right border sequences

2) Virulence Region (~40kb): All genes whose products are necessary for
transfer of T-DNA from Agro plant cell.....
• Formation of T-DNA intermediate
• Formation of channel from Agro to plant
• Shuttling of T-DNA thru channel to plant cell
• Nuclear targeting of T-DNA
• Chromosomal integration

3) Genes for synthesis of opines (serve as source of C &
N for Agro)

phenolic derivatives (eg. acetosyringone) released by wounded plant cells
- chemoattractant
- induce vir gene expression
- turns on host replication and repair machinery

Modifications to Ti Plasmid to Make it A Useful Plant
Transformation Vector
1. Removal of Onc genes from T-DNA (keeping border repeat
sequences intact)

2. Replacement of Onc Genes with Plant Selectable Marker Gene (kanamycin resistance)

3. Insertion of convenient Multicloning Site
For easy insertion of your gene of interest
between the T-DNA border repeat
sequences

4. Placement of Vir Gene Region (of Ti plasmid) on
a Completely Separate Plasmid
The Vir genes still work to transfer the (now modified)
T-DNA region to plant

I. Agrobacterium-mediated transformation (stable transformation only for dictos and gymnosperms):
Agro cells (containing 2 plasmids – one with T-DNA, the other with the Vir
region) are used to infect wounded plant cells (eg. leaf discs)
Leaf discs placed on medium to induce shooting, then rooting (all in presence
of selectable marker – Kan)
Kan-Resistant plantlets generated screen for expression of gene of interest



II. Direct DNA Transfer Methods (Alternatives to
Agro-Mediated Transformation) (these are less preferred, more destructive)
A. Microprojectiles (Gene Gun Method or Biolistics) - DNA coats surface of tungsten or gold particles
B. Microinjection: - use of holding pipette to deliver DNA solution directly into the cell
C. Electroporation - DNA taken up by recipient plant cells; electric
pulse is used to generate transient pores in plant
plasma membrane

With all these methods:
Assay for gene expression after 24-48 h
Go for stable transformation (attempt to
regenerate whole plant)

III. Applications: Engineering pathogen (microbes, bacteria, fungus, virus) resistance
-------------------------------------------
I. Plant diseases caused by microbial pathogens

How do pathogens reduce crop yields?
Cause tissue lesions
Reduce leaf, root or seed growth
Clog vascular tissues and causing wilt
Cause general metabolic drain, in the absence of external
signs of damage
Cause pre- or postharvest damage (blemishing total decay)

What factors cause crop devastation?
The Disease Triangle:
(1) Pathogen (genotype & prevalence or mode of introduction): Virulent
pathogen must be present in sufficient numbers at the right place and time
to start off the epidemic
(2) Plant (genotype and planting configuration): susceptible plant varieties
must be present
(3) Environment (pathogens are sensitive to temperature, humidity, wind
and weather conditions)

Two major factors can contribute to devastation:
Monoculture & Genetic Uniformity
1. Monoculture: Growth of a single
crop spp. on a large piece of land;
strong regional emphasis on a
given crop

2. Genetic Uniformity
Permitting genetic uniformity - the more dangerous practice
Farmers gravitate toward the most successful varieties of a crop
in terms of yield..... tendency to use fewer and fewer plant
genotypes

The best strategy is to maintain genetic diversity among the
different popular varieties of a given crop spp.
Disadvantages:
• non-uniform crop
• mechanical harvesting is not an option

Plant Viruses - potato virus x (rna), coat proteins
Virus symptoms
• lessions (spots) on leaves
• Can include a mosaic patterned
yellowing of leaves
• Leaf distortion & curling
• Raised bumps & mottling of squash
fruit

Bacterial wilt disease of cucumber, Xylella fastidiosa invades xylem, blocking water transport, causing wilting

Plant Fungal Diseases
Maize ear rot - caused by fungi that produce mycotoxins (harmful to humans and animals)

Wheat rust (fungus) diseases (stem & leaf)
- Rusts are the most destructive plant diseases known

Identify types of genes involved in pathogenicity of fungus:
proteins for synthesis & secretion of toxins
enzymes to break down plant cell wall
enzymes to detoxify plant defense chemicals
sugar transporters - support existence in nutrient-poor
xylem sap
regulatory proteins - adjust gene expression for growth in
different environments
synthesis & secretion of extracellular polysaccharides (slime)
efflux of antibiotics (produced by plant to kill bacteria)
uptake & sequestration of iron and other metals
Why is this important?
Devising novel approaches for control of pathogen

How do fungi enter the plant and cause disease?
*Enter at wound sites in the plant
*Secrete enzymes that hydrolyze plant cell walls
*If plant defense responses are not sufficiently
rapid, fungal hyphae quickly grow and spread
from cell to cell

Once fungal pathogen is inside plant cell......
- Some produce toxins alter permeability of membrane
- Some secrete slime accumulates in vascular tissues
wilting/death due to lack of transport (water & nutrients) from
roots to shoots
- Some produce plant hormones (plant loses control over
organized growth & development)
- Some attack seedlings as soon as they emerge or even before, eg.
soil-borne fungi (“damping-off”)
- Many cause necrotic (dead) spots on leaves, stems, fruits,
& seeds decrease vigor of plant; render seeds & fruits
less fit for human consumption

chemical strategies for disease control:
- Fungicides – applied as a seed coating prior to sowing. Or, as
sprays or as dusts on plants in field.
- Antibacterial - Copper or sulfur sprays & antibiotics

Problems with chemical strategies for disease control
1. Expensive to use
Impractical for grain crops (used on vegetable, fruit & flower crops)
2. Repeated use often leads to pathogen resistance
Some bacteria have genes encoding proteins that allow them to degrade,
export or otherwise resist the compound
Applies strong selection pressure ...
An initially small population of resistant bacteria becomes dominant in
population within the area
3. Human toxicity or broader environmental toxicity to non-target organisms

II. The biology of plant-pathogen interactions

Three general principles:
1. Plants defend themselves by using preformed defenses
(constitutive) & by turning on (inducible) defense genes
2. The key to plant resistance is swift induction of defense-related
genes ... in turn depends on early plant recognition of pathogen
3. Successful pathogens elude plant defenses

(1) Defenses always in place (“constitutive”) / passive:
1. Thick cell walls & waxy cuticle on surface of leaf & stem (wax dries out
rapidly less support for growth of fungi & bacteria)
2. Plants produce diverse array of antimicrobial compounds (mostly active/inducible)
Preformed inhibitors (glucosides, saponins, alkaloids)
Antifungal proteins
Antifeedants
Enzyme inhibitors
Not all are constitutively produced; some are induced....
Plant does not want to devote its metabolism to defense

Response of plant to microbial infection is multifaceted:
* Increased ("up-regulated") expression of a large number of pathogenesis-related genes in cells at site of infection (eg. chintinase and glucanase degrade cell walls of invading pathogens), antimicrobial (thionins, defensins, lectins, phytoalexins)
* Activation of pre-existing enzymes that control synthesis of anti-microbial compounds
* Strengthening and cross-linking of the cell walls
* Secretion of phenolics (eg. lignin - dense phenolic – based
polymer (network), eg. salicylic acid - precursor to aspirin for anti-inflammatory) into the cell walls
* Generation of signaling molecules move locally or systemically to activate defenses in other plant cells: systemic acquired resistance
* In some cases, the hypersensitive response (seen as necrotic spots).... a beneficial plant cell death response cells immediately surrounding the infection site die, effectively preventing spread of pathogen

Summary of different types of plant defense


(2) Plant recognition of Pathogens
Plant must be able to recognize presence of a pathogen.....

Recognition: binding of molecule derived from pathogen ('elicitor') to molecule (receptor) (R-protein) of the plant

Resistance genes (R-genes): Plant genes encoding recognition (receptor) proteins

Each R-gene encodes a protein that recognizes a specific pathogen compound activates host defense responses

Elicitor may be:
• a virus coat protein eg. Tobacco mosaic virus
• a bacterial virulence factor (secreted into host plant)
• a fungal protein present on pathogen surface
• cell wall fragment?

When strong resistance defense responses are elicited in plant -> formation of necrotic spots
Represent infected plant cells that plant has actively sacrificed to prevent spread of pathogen
Rapid hypersensitive response is a programmed cell death process... it is adaptive because:
• it effectively “walls off” the pathogen
• releases antimicrobial compounds
• releases signaling molecules that elicit defense responses in
other host cells: systemic acquired resistance
• kills off host cells that might otherwise support the growth
of the pathogen (virus/fungus)

How do pathogens successfully evade plant recognition?
- Different R gene products control defense activation: detect extremely different pathogens (viruses, bacteria, fungi, nematodes or insects).
- Proteins encoded by R genes share similar structures & mechanisms for pathogen recognition are highly conserved across different plant species and diseases

Why is this significant??
- different specifics of pathogen recognition evolved from small number of progenitor R genes
- New R genes with new pathogen recognition capabilities arise over time. This evolution has been crucial for the ongoing battle of plants to keep pathogens at bay.

- Over-use of chemicals to combat pathogens -> Strong selection pressure resistant population emerges
- Same thing occurs when pathogens face R genes
- Elicitors .... Usually part of the pathogen structure
- Pathogen evolves to elude the plant’s recognition system

Good news is that plants have around 100 R genes

III. New biotechnological approaches to create plants with
enhanced disease resistance

Classical crop protection strategies (plant breeding) -> id. R genes in wild plants & older crop var.
-Of same or very closely related species
Forest industry: Rust-resistant pine

With a genetic engineering strategy...
R genes (& other genes inv. in defense) can be isolated from one plant spp. and introduced into another
No need for sexual compatibility

What are some major genetic engineering strategies?
(1) Expression of genes encoding specific antimicrobial compounds
(e.g. PR proteins)
· Hydrolytic enzymes (chitinases - Hydrolyze b-1,4-linkages within chitin polymers of fungal cell wall, glucanases
· Antifungal proteins (osmotin- and thaumatin-like)
· Antimicrobial peptides (defensins, lectins, lysozyme)
· Ribosome-inactivating proteins (RIPs)
· Phytoalexins
(2) Expression of genes encoding products that can potentially
enhance the structural defenses of the plant
· elevated levels of lignin
(3) Expression of genes encoding products that destroy or neutralize
a component of the pathogen arsenal
· gene for oxalate oxidase, involved in the degradation of oxalic
acid)
oxalic acid –inactivate plant cell defense enzymes?
(4) Expression of genes encoding products that result in the release of
signals capable of regulating plant defenses
· specific elicitors
· H2O2
· salicylic acid (SA)
· ethylene (C2H4)
(5) Expression of genes encoding defense-activating “master switch”
proteins



(1) Expression of genes encoding specific antimicrobial compounds
(e.g. PR proteins)
· Chitinases can defend against fungal attack & invading fungal hyphae
· Chitinases: basic or acidic - acidic forms are extracellular; basic forms are found in vacuole

Problem: In many cases, only partial resistance is obtained

npt = neomycin phospotransferase (plant selectable marker gene Kan-res)

But! Researchers are expressing different combinations of genes:
greater resistance is the outcome

Eg. Expression of genes encoding a chitinase & b-1,3-glucanase (tomatoes)

Eg. Expression of chitinases & RIP (ribosome-inactivating protein) in barley


(4) Expression of genes encoding products that result in the
release of signals capable of regulating plant defenses

eg. salicylic acid (SA) will activate SAR when sprayed on plant -> system primed for defense

eg. Expression of elicitor in plant (viral coat protein), interference of viral protein with viral RNA replication, viral movement from cell to cell


(5) Expression of genes encoding defense-activating “master switch” proteins

Expression of R gene products (eg. constitutive Cf9 R protein) involved in HR and in interaction with avirulence (lack of virulence; lack of competence of an infectious agent to produce pathologic effects) (inducible Avr9 elicitor) factors => hypersensitive defense.


Plant defense against biotic stress - pests (insects, herbivores, cows, worms, weevil, flies, etc.)
-------------------------------------------
Plants are surrounded by hungry herbivores
– Herbivores range in size from microbes to cows, white pine weevil, Cotton boll weevil, Cotton bollworm (eats developing fruits), Caterpillars, army worm, pine bark beetle, Potato beetle, whitefly (sucks phloem sap),

nonnative plant pest - pest migrate to a new country without their predators

Predator populations – determine insect pest populations

Chemical defenses:
- Natural products:
a) terpenes
b) phenolic compounds
c) nitrogen-containing secondary product

- Plant secondary metabolites include chemicals we use as
• Drugs (medicinal and recreational)
• Dyes
• Perfumes
• Beverage manufacture
• Poisons



(a) - Terpenes/Terpenoids - mono-, sesqui-, di-terpines, insect repellant, feeding deterrant, oleoresin (both constitutive and induced) in conifers
-------------------------------------------
• Based on 5C isoprenoid unit:
- Monoterpenes 2 x 5C
- Sesquiterpenes 3 x 5C
- Diterpenes 4 x 5C
etc.

• Roles:
– Plant growth and development
• Gibberellins are diterpenoids 4x5C, ABA is a sesquiterpenoid - 3x5C (hormones)
• Carotenoid pigments are tetraterpenoids 2x(4x5C)=C40
– Photosynthetic pigments & protect against high light
– Plant defense
• Toxic and/or feeding deterrents for herbivores (monoterpenes - 2x5C) - insect repellants and triterpenes - 2x(3x5c) - feeding deterrants
- Phytoecdysones disrupts molting (also known as sloughing, shedding or for some species,)

Defence against insect herbivory
• Responses to insect herbivory involve the wound response
– Lead to inducible chemical defenses:
• affect attacking insects
• affect natural enemies of attacking insects
• BUT most resistance is “constitutive”
– Results from pre-existing chemical or morphological
defenses

Terpenoids and insect defence in conifers:
The major defense against insect and pathogen
attack in conifers is the oleoresin
– Complex mixture of mono-, di-, and sesquiterpenes
– Mono- and sesquiterpenes are volatile and provide
fluidity to resin
• Resin (Resin is a hydrocarbon secretion of many plants, particularly coniferous trees, used in nail polish, fossilizing/amber insects http://en.wikipedia.org/wiki/Resin) flows to point of injury
• Insects are exposed to toxic terpenoid components
– Diterpenes seal wounds

Resin defenses are constitutive and induced
– Constitutive defenses (pre-formed traits)
• First line of defense, repel attack
– Induced defenses
• Second line of defense
• Resin composition differs
– More insect toxic?

The white pine weevil is a pest of
regenerating Sitka spruce

Weevil attack induces the formation of induced
resin canals in the xylem

(b) Phenolics - flavonoids (anthocyanin-color, flavonol-color in flower, isoflavonoids-anti-cancer in legumes, anti-estrogen, tannins-feeding repellants, non-specific protein binding=toxic, wine), lignin=structural,
-------------------------------------------
Plant phenolics are biosynthesized in several different ways. In higher plants, most phenolics are derived at least in part from phenylalanine, a product of the shikimic acid pathway

1) Lignin
- Dense polymer made up of network of phenolic units
- Provides mechanical support to plant
- Synthesis is induced by pathogen infection or insect wounding

Lignin resists attack by most microorganisms. Lignin is nature's cement along with hemicellulose to exploit the strength of cellulose while conferring flexibility.

2) Flavonoids:
- Basic structure: 15 Cs arranged in 2 aromatic rings connected
with a 3C bridge
4 groups:
• anthocyanins - colored flavonoids that attract animals
• flavones and flavonols - Flavonoids of flowers, attract bees and N2 fixers, short wavelength
• isoflavonoids - one aromatic ring is shifted, act as anti-estrogens => infertility, anti-cancer
• Tannins: condensed (feeding repellants, toxic) or hydrolyzable (gallic acid - antioxidant, red wine),

Anthocyanins - The structures of anthocyanidins (A) and anthocyanins (B). The colors of anthocyanidins depend in part on the
substituents attached to ring B (see table). An increase in the number of hydroxyl groups shifts absorption to a
longer wavelength and gives a bluer color. Replacement of a hydroxyl group with a methoxyl group (OCH3) shifts
absorption to a slightly shorter wavelength, resulting in a redder color.

Isoflavonoids (Isoflavones):
• found in legumes
• some are insecticidal
• some act as anti-estrogens:
- sheep grazing on clover rich in isoflavonoids can
suffer from infertility
- anti-cancer benefits of soy-based foods

Tannins (condensed/ hydrolysable - gallic acid):
• Polymerization of flavonoid units
• Condensed tannins found in seed coats of legumes:
toxic towards some seed-eating beetles
• Significantly reduce growth of many herbivores
• Feeding repellents for many animals ( eg. unripe fruits
with high tannin levels avoided by deer and cattle)

The defensive properties of tannins are due to their toxicity
- ability to bind proteins non-specifically

Tannins:
• Red wine polyphenolics (tannins) have health
benefits:
- block formation of endothelin – 1, a signaling molecule that makes blood vessels constrict
- benefits for heart disease

(c) Nitrogen – containing secondary compounds (alkaloids-cocaine, cyanogenic glycosides-HCN gas-found in seeds of almonds etc., aa-analogs-canavanine):
-------------------------------------------
Alkaloids (nicotine, cocaine, morphine, codeine- for analgesic(pain relief), etc):
• These can be extremely toxic
• Synthesized from amino acids (terpene pathway supplies C – skeleton)

Alkaloids:
• Very effective deterring insect attack
• BUT!! Place strong selective pressure on predatory insects
to overcome defense mechanism
• Some herbivores (oxidizing alkaloid instead of reducing it in the gut) can become adapted to tolerate one class
of alkaloids

Cyanogenic glycosides:
• When broken down, release poison – hydrogen cyanide (gas)
• Not broken down in intact plant
• Leaf damage due to herbivore feeding allows hydrolysis
• Some CGs are found in seeds of almonds, apricots, cherries,
peaches, etc.
• HCN – toxin that inhibits metalloproteins (eg. cytochrome oxidase)
• Tubers of cassava

Non-protein amino acids:
• Play a protective role in some seeds
• Eg. amino acid analogue (canavanine – an analogue of Arg)
is produced in seeds of Brazilian vine
• Toxic to most animals & insects: inability to distinguish analogue
from aa Arg vs. non-functional protein (3-D structure or catalytic
site is altered)


Other defensive proteins
a. Inhibit herbivore digestion: - α – amylase inhibitors, Proteinase inhibitors, Lectins
b. Inhibit protein synthesis:- Ribosome–inactivating protein (RIPs)

Enzyme inhibitors:
• Insects use amylases and proteases to digest the
starch & protein in their food
• Many seeds (especially legumes) contain inhibitors
of insect digestive enzymes:

Protease inhibitors (PIs) - Inhibitors of insect proteases such as trypsin, chymotrypsin, elastase & subtilisin

Mechanism of inhibition - inhibitor forms a strong
covalent bond with the active
site of the insect protease

• Insect larvae starve to death due to loss of nutrients
& over-production of proteases

Protease inhibitors:
Leaves of various plant species (e.g. tomato, potato)
rapidly synthesize PIs
• in response to mechanical damage (insect attack/
wounding)
• synthesis occurs throughout the plant
• stored in central vacuole as defense against
repeated attack

alpha-amylase inhibitors:
Unlike peas, beans are not attacked by pea weevils.
Beans contain a protein that inhibits the activity of
alpha-amylase, an enzyme that helps in digestion
of starch. This protein inhibitor, called α-amylase
inhibitor, causes the weevils feeding on beans to
starve before they cause any damage.

Lectins:
• Carbohydrate-binding proteins (in plant tissues, seeds):
different lectins have different sugar specificities
• After ingestion by an herbivore, lectins bind to epithelial
cells lining digestive tract (interfere with nutrient
absorption)
• up to 30% of total protein in some seeds
• lectin from bean: toxic to developing larvae of bruchid
beetle (may bind to midgut ephithelial cells)


Ribosome-Inactivating Proteins (RIPs):
“ Jack in the box”
- e.g., highly toxic ricin found in castor bean seeds
- contain a lectin chain linked to a polypeptide that inactivates
ribosomes (hydrolyzes the sugar base linkage at one specific
position in the rRNA)
- highly toxic towards Coleoptera (e.g., boll weevil and bruchid
beetle) and locusts
- not good candidates for genetic engineering
“ Bulgarian Diplomat “


Combating insects with chemical insecticides:
Chemical insecticides (e.g., organochlorines – DDT):
Problems:
(i) Toxicity (to non-target organisms – pollinating insects,
natural predators of pests, humans)
(ii) Environmental spread (more than 99.9% is wasted;
substantial economic cost)
(iii) Loss of effectiveness (build up of populations of
resistant pest species) – high selection pressure
imposed by toxicity and heavy application
(iv) Consumer pressure – public concern over pesticide
residues in food stuffs


Biological control:
- Use of predatory or parasitic insects, nematodes
and fungi (eg. aphid parasite (parasitic wasp) on greenbug aphids)
- Effective in confined areas; success in field is
limited
- Requires that a population of
the pest has built up and that
the biological control organism
is not itself competed out


Strategies for genetic engineering of plants
for insect resistance

Transgenic plants expressing:
Protease inhibitors
α-Amylase inhibitors
Plants have evolved amylase and protease inhibitors to specifically
inhibit the digestive enzymes of certain insects
Transgenic tobacco expressing cowpea trypsin inhibitor (non-toxic to humans):

Differences in organization of mammalian & insect gut:
- in mammals: any inhibitor would be exposed to acid pH 2 of stomach where it is exposed to pepsin first, then Digestive enzymes (trypsin, chymotrypsin) in small intestine ph8
- insects don't have pepsin before passing to mid-gut ph9-11

Transgenic plants expressing genes for α-Amylase inhibitors:
• Amylase inhibitor of kidney bean – completely inhibits growth of pea
weevil and cowpea weevil
• Introduced bean gene into peas (got high level expression):
development of pea weevil larvae was completely inhibited
• BUT! Animal feeding experiments show deleterious effect of transgenic
peas


Plant defenses against insect herbivores:
1) Constitutive defense responses (mostly morphological - leaf hair):
- always present
- species-specific
- stored in less-damaging form
“Constitutive Defenses”
Eg. chemicals constitutively
produced by leaf hairs trap
and kill larvae

2) Induced defense responses:
- initiated only after actual damage occurs
- same defense chemical may be involved in
constitutive and inducible responses


Insect herbivores can be classed by degree of
damage inflicted on plant:
Least
damage
1) Phloem feeders
2) Cell content feeders
3) Chewing insects
Most
damage

1) Phloem feeders:
• Aphids and Whiteflies
• Direct injury to plant
is low, but insect may
vector plant viruses:
aphids spread barley
yellow dwarf virus, a
common disease of
cereals

2) Cell content feeders
• Mites and thrips
• pierce plant tissue
• intermediate damage


3) Chewing insects
• caterpillars
• grasshoppers
• beetles
• Cause significant damage to plant
• Can vector fungal pathogens
Responses to
insect herbivory
and pathogen
attack overlap


Insect defence in angiosperms
• Herbivory injury is often mimicked in part by wounding
• Leaf wounding causes Systemic response in
distal, unwounded leaves
– Local responses
– Systemic responses
• Rapid
Local response in wounded leaf
• Insect attack/wounding results in mobile signals emanating
from damaged tissue
• As well as a wounding response, the plant may recognize
insect – derived compounds: “elicitors”

Outcome of local and systemic responses:
1. Direct defense responses
• Production of proteinase inhibitors (PIs) + other “nasties”
• Decrease palatability of plant or fitness of insect
2. Indirect responses
• Production of volatile organic chemicals (VOC)
• Target predators or parasitoids of attacking insect

Elicitors present in insect saliva:
• Fatty acid – amino acid compounds
• Ingested plant tissue supplies source of
fatty acid (eg. linolenic acid in plants, 18:3)
• Enzyme in gut of insect conjugates plant
FA to insect amino acid Gln
• When plant recognizes elicitors present
in insect saliva – signal transduction
pathway is triggered:
increases jasmonic acid (JA)

Insect defence in tomato
• PI production is used as a ‘marker’ for wound responses
& can be induced by the following signals:
– Oligogalacturonides (OGA)
• Cell wall fragments released due to damage/enzymatic
degradation
– Jasmonic acid (JA)
• Lipid-derived (oxylipin) signalling molecule (derived from
linolenic acid)
– Systemin
• Peptide (18 amino acids long) produced due to proteolytic
cleavage of a precursor polypeptide (first peptide hormone
identified in plants)
• Mobile signal?
• Triggered by wounding and insect herbivory


wound > prosystemin > systemin > LRR receptor > PLA2 > JA biosynthesis > JA translocated via phloem to target cell to encode protease inhibitor

Insect elicitors modify the wound response: tobacco hornworm M. sexta (normally blue, but eats yellow carotenoid and so it turns green) attacking tobacco plants that germinate in response to wood smoke

Interestingly, the nicotine in the leaf is normally toxic, but the caterpillars have a mechanism for selectively sequestering and
secreting the nicotine.

Insect elicitors modify the wound response
• Nicotiana attenuata produces nicotine as a major chemical defense
– Nicotine poisons acetyl choline receptors at nerve-muscle junctions
• Nicotine synthesis is induced by wounding
• When the plant is attacked by nicotine tolerant M. sexta, there is a
decreased production of nicotine
- pest detoxifies/sequesters nicotine
- pest modifies the plant wound response
• Instead, volatile terpenes are released
- attract insect predators of the pest
- decrease oviposition of adult moth

M. sexta oral secretions and regurgitants are
sufficient and necessary to modify the wound
response
– Fatty acid-amino acid conjugates


Insect elicitors modify the wound response



Who benefits?
– Insect?
• Less nicotine produced
• Reduces growth penalty assoc. with detoxification
– Plant?
• Optimizes indirect defenses
• Reduces costly metabolism directed towards nicotine
production
• Other direct defenses are not affected
– Slows development of M. sexta
– Increases opportunity for predation

Lepidoptera – (worms) Caterpillars, corn borers,tobaccobudworm/hornworm, corn earworm, army worm
Coleoptera–Cotton boll weevil, bruchidbeetle
Others –White pine weevil, cone/seed eating pests, pine bark beetle

toronto

http://www.blogto.com/
http://spacing.ca/wire/

Seattle Freeze?
http://seattletimes.nwsource.com/pacificnw/2005/0213/cover.html

"On the one hand, it's nice to bop in and out of situations knowing people will smile and treat you well. Nice is like bubble gum — it's sugary and pleasant." But if all you ever get is nice, never flirty or risky, she says, that gum loses its flavor pretty quick, and the human experience becomes ultimately less rewarding. Even depressing.

"I like your bag"
A random coupling story that always made me smile was my friend that met his soon to be girlfriend on the streetcar. It was rainy they were both soaked and tired from a long day of work and there were two seats left at the back of the car. They politely walked into each other twice doing that, "After you." "No I insist, after you." thing we do when we're embarrassed and just sat beside each other. My friend smiled at her and said, "I like your bag." BAM! That was it. Seriously. They exchanged numbers a minute or so later and were dating for four years.


It seems like the harder you look for it the harder it is to find.

http://www.yelp.ca/

life

hmm ... things starting to go haywire, got my bisc mark back lower than the avg, i don't really like how that course is marked but oh wellz ....

two more exams to go and i'm out of undergrad for good ... don't know what'll happen next month when i'm done ...

a review of what happened in the past so far ... lets c (in case i become demented)

..::<[the chocolate]>::..
crush when i was a kid, flew to another side of the world, later met, gave chocolate on x'mass, talked on msn
... didn't last very long, distance thing, never really responded back, got married this year of 2010 ...

..::<[the sub]>::..
one day, hear the person's forgot their lunch, commuted all the way from school, skipped tutorial (which i rarely do), bus hopped a couple of times, finally met up at the bus stop, got a smile after giving the sub and h20 ... drew a pic, slided it in a big blue binder, the person probably didn't see it ... oh wells, invited the person to tennis, invited last week for a walk too, went to halloween on invitation of a friend though but that was it ... also getting married this year of 2010 ....

..::<[the egg pie]>::..
hmmm ... didn't really go far compared to the others, had in one time, touched hands like the one in tarzan, went after work to celebrate birthday, delivered food to school, i had a feeling this person was using me to get food, how sad .... well, the person's in medschool now, maybe happy with someone else ...

..::<[the compass]>::..
met at a class by chance, going to a familiar place in the summer, did assign together, and always get lost going to a place, so gave the person a compass,
asked the person twice for a walk, declined, ..... seemed to ignore/avoid too, .... life's harsh ... probably for the best anyways

..::<[the guide]>::..
the person actually agreed to meet after asking twice, went for dinner (waited for 30 min cuz i said i was gonna be late but was actually on time, my bad), gave a pic too, temple, karaoke, had a feeling the person wasn't happy, was it only as a guide?

..::<[the adventurer]>::..
this person seems to be the only one that came closest, the person is more responsive than the rest, but alas, the distance thing kicks in .... tough ... tough

..::<[the end]>::..
and so that concludes these few years of set backs, this thing is meant for some people, not mean for others, perhaps not enough effort was put in, perhaps didn't read the signals correctly, perhaps not really a 'people' person, too sensitive?

but hey, a person can feel something too, it's probably for the best, get out of the hole while it's still shallow before falling into an abyss where escape would be an epic endeavor.
from now on, try to go fwd, only look back so that these lessons are not takn 4granted.


http://www.flickr.com/photos/bazstyle/2274293698/
"Every cloud has a silver lining." http://www.usingenglish.com/reference/idioms/every+cloud+has+a+silver+lining.html

finally took a load off my chest, maybe i'll regret writing this later, maybe not, oh wells oh wells oh wells

Mobile phones and social relationships

http://blog.dialaphone-blog.co.uk/blog/2008/04/23/10-ways-a-cell-phone-can-ruin-your-relationship/

lol.


on the bright side ...

http://www.healthguidance.org/entry/11764/1/Do-Mobile-Phones-Improve-Relationships.html
"The extent of the phone us was irrelevant but the students felt more social support if they had more contacts. The number of contacts also impacted positively on the number of calls and texts received and texts sent.

It seems then that just as guns don’t kill people (people kill people), mobile phones neither ruin nor improve relationships on their own – this is achieved just fine by the people on either end. "

http://www.smh.com.au/news/phones--pdas/mobiles-key-to-modern-relationships/2007/07/16/1184438206905.html
"according to an Australian-first study, the mobile phone is a crucial aid in modern relationships."

... interesting ...

what if's

youtube streaming music

fb doodle over pic

hypersensitive response (program cell-death - apoptosis, around infection, forming necrotic spots) in plants work with humans?

rice and weeds

armor / jacket made of lignin

http://en.wikipedia.org/wiki/Lectin

Predator populations – determine insect pest populations

Isoprene (short for isoterpene[1]), or 2-methyl-1,3-butadiene, is a common organic compound with the formula CH2=C(CH3)CH=CH2. It is present under standard conditions as a colorless liquid. It is the monomer of natural rubber and is a precursor to an immense variety of other naturally occurring compounds.
http://en.wikipedia.org/wiki/Isoprene

Poisons/Toxic - Nitrogen – containing secondary compounds: Alakloids-cocaine, Cyanogenic-glycosides, aa-analogs (cananavine)

Week of Music

Spontania feat.JUJU / 君のすべてに
http://www.youtube.com/watch?v=hdDI1TAydqc&feature=related

JUJU 明日がくるなら 　PV
http://www.youtube.com/watch?v=vdlo3OiLaUk&feature=related

Jasmine ジャスミン- Sad to Say
http://www.youtube.com/watch?v=UG6q4O8Wl8A&feature=related

http://www.4shared.com/dir/23977917/73ed277d/japanese_songs.html

Forensics

forensic biology:
--------------------------------------
- hair
- body fluids: blood, semen, urine, saliva
- DNA

hair: 1. cuticle (covering) 2. cortex (color) 3. medulla (canal)

sexual assault kit:
- used to collect evidence
- contains swabs, pubic comb

presumptive test - cheap, quick test, covers larger areas, have false positives

confirmatory test - more expensive, no false positives

blood:
- presumptive test - haemastix test, luminol
- confirmatory test - Haemochromagen chemical test, DNA typing

semen:
- presumptive test - fast blue test (phosphate color test)
- confirmatory test - microscope, prostate specific antigen (PSA)- relies on semen, not sperm, DNA typing
- must be collected within 48 hrs, if later, sperm is destroyed

DNA:
- every indiv. has dif’t lengths of tandem repeats
- when 2 samples compared, several loci known to have STR’s (Short Tandem Repeats) will be examined to increase validity of results
- restriction fragment length polymorphisms (rflp) - expensive, slow 6-8 weeks, need high quality DNA
- polymerase chain reaction (pcr) - inexpensive, fast 1-2 weeks, may also amplify contaminants
- database on frequencies of STR used to see if results are statistically significant
- Bill C-104 allows DNA sample to be collected under warrant
- buccal swab - non-invasive method of collecting DNA by rubbing swab inside cheek of the mouth
- mtDNA - mitochondrial DNA - inherited from mother alone, more useful in identifying victim than suspect
- egg - provides genetic material, cytoplasm & all other cell materials

forensic chemistry:
--------------------------------------
- paint, glass chips, fibres, bomb, arson, drywall, glues, cleaners & tape
- qualitative and quantitative (% composition)
- mass spectromphotometry: characterize & identify many purified substances
- gas chromatography: (purifies) separates components of a mixture of substances & tentatively identifies them

Significance of evidence:
- explain in jury in terms of significance
- probability that evidence will be found on someone else
- depends on circumstances

fibre - color, refractive index, infrared spectrophotometry results

paint - many layers, # of variants determine year, make, model

fire - any accelerants?

bomb - how was it constructed, composition, detonating mechanism, Explosive ² timers, detonators, fuses, batteries, something to hold it together ² duct tape commonly used

explosion - fast immense pressure (expand & compress layers of surrounding air
) ² stretches & explodes rigid pipe producing sharpnel (sharp, deadly fragments)

types of explosive (according to their rate of decomposition):
i. low explosive - decomposes slowly "deflagration", used for propelling rockets in a specific direction
ii. high explosive - usually nitration based, decomposes quickly "detonation" due to shock
a. primary explosive - ultrasensitive to heat shock, detonates easily and violently
b. secondary explosive - insensitive to shock, usually detonated during initial explosion ie dynamite, TNT

forensic toxicology
--------------------------------------
1. deals with detection of drugs or poisons
2. involves interpreting findings - physiological/behavioral effects
determine WHAT DRUG it is and HOW MUCH of it is present & WHAT BEHAVIOUR a person would make

- drug metabolites deposited in hair
- assist in establishing true cause of death
- “the dose makes the poison” - anything can be a poison, depending on dose, ie. oxygen & water = too much or too little can cause death

2 main parts of forensic toxicology investigation:
1. analytical phase - specimen (eg. blood, urine) is analyzed for drug's metabolites, smaller chemical substances to identify the drug, gas chrom. & mass spectro
2. interpretive phase - when drug was last taken, amount, what symptoms, chronic (drug taken over long period) or acute (sudden, one-time exposure)

specimens:
blood - easy to get, primarily used for drug screening (confirming and quantification), but it needs to be purified 'dirty' speciment, invasive, may not be available after body has decomposed, decomposition of proteins by bacteria (putrefaction) produces alcohol in blood (must take blood samples from diff. parts of the body)

urine - second most popular for substance analysis (confirmation only), clean specimen, less invasive than blood

hair - determine drug history, not invasive, only from criminal cases in Canada

virtuous humor (eye) - protected from putrefaction

other other non-biological exhibits like surrounding syringe, spoons, pipes, medication, food

presumptive tests - narrow down to a class of drugs
confirmatory tests - gas chrom. & mass spectro

clandestine lab - produces illegal drugs, dangerous so forensic chemists help police identify danger and identify what they want to collect for evidence

forensic toolmark
--------------------------------------
- determine whether tool mark was produced by particular item
- physical matching: usually firearms or tool marks, but also physically matching glass from a hit & run or broken scissors or knife tips in a wound
- tool: any hard object that makes an impression in a softer object (ie. Chisels, hammers, tire impressions in snow, footprint in blood)
- restore serial #
- attend autopsies to reconstruct event, entry or exit wound, how far was the shot was fired

2 types of tool marks:
1. impressed - no movement of tool other than force or blow
2. striated - pressed with sliding motion

class characteristics - eg width and shape of knife blade, used for elimination of tools
individual/accidental characteristics - formed during manufacturing or use,

examiner makes impression of suspect tool (in very soft substance), which can then be compared w/ suspect mark directly

some cases, similarities btw. tool marks from dif’t tools
- dif’cs btw. tool marks made by same tool

causes of false similarities: same manufacture carryovers (dirt, imperfections) between tools, by chance, unfamiliar tool

causes of false differences: same tool corrodes over time, over use, diff't softness of objects, tool deliberately changed, cleaning

serial # restoration:
- strained areas will dissolve faster (using etching agent eg acid) than unstrained metal & # will reappear, doesn't work for laser writing because they don't strain metal


Forensic Firearms
--------------------------------------
- cartridge loaded in gun has many parts ( metal case contains propellants, gunpowder, bullet) (*gunpowder needs small charge to set it off - tiny granules)
base of cartridge usually contains primer ² ignites smokeless powder & ejects bullet when struck (see http://en.wikipedia.org/wiki/Bullet)

A modern cartridge consists of the following:
1. the bullet itself, which serves as the projectile;
2. the case, which holds all parts together;
3. the propellant, for example gunpowder or cordite;
4. the rim, part of the casing used for loading;
5. the primer, which ignites the propellant.
- bullet (projectile or “slug”): piece of cartridge shot out of gun

bullet types:
1. full metal jacket - used in military, disables target, prevents from fragmenting
2. hollow point - mushrooms, used in hunting,
- A hollow point is an expanding bullet that has a pit or hollowed out shape in its tip, generally intended to cause the bullet to expand upon entering a target in order to decrease penetration and disrupt more tissue as it travels through the target

http://en.wikipedia.org/wiki/Full_Metal_Jacket_bullet
Whereas hollow point and soft-tipped bullets are designed to expand upon impact, fully metal jacketed bullets are technically limited in mechanisms to increase round expansion. In some cases this leads to smaller target damage, although not in all instances.

cocking - when cocked, ready to be fired
² moves hammer towards webbing btw. thumb & forefinger
² trigger releases hammer & it falls rapidly, striking primer
² bullet (softer object) travels @ great speed along hard barrel (tool)

barrel drilled again to produce rifling: series of spiral grooves (if straight barrel, tumble over end & not fly in straight line)

dif’c in # of grooves cut, width btw. them, whether spin bullet clockwise or anticlockwise

diameter of original barrel, before rifling, determines caliber of gun (class characteristic)

* most test bullets in large water tank ² water slows bullet rapidly

test bullet & suspect bullet compared directly w/ suspect bullet using bullet holder underneath one objective of a comparison microscope

- hammer = tool (hard object) ² leaves mark on softer object, primer

like in tools, similarities btw. bullets fired by dif’t guns or difc’s in bullets fired by same gun

diff. when using diff types of bullets not designed for the gun, cleaning, cheap guns

gunshot residue (powder never totally burns):
- gun fired ² leaves gunpowder residue behind
- some gun powerder flashes back @ shooter

gunshot residue analysis - determine how far (distance) shooter was from target who fired shot

distance of gun from target determined by: - amt - type - pattern of residue

Types of shots:
1. contact shots - nearly all unburned powder is forced directly into wound, skin around bullet wound tears in star shape pattern
http://www.ojp.usdoj.gov/nij/training/firearms-training/module12/fir_m12_t06_02.htm

2. Close distance shot (less than 1 inch): bullet wound has rim of heavy concentration of smoke-like vaporous lead, clothing is scorched,

3. Shot from 12 - 18 inches: soot (incomplete combustion of carbon) & powder deposited around bullet hole together

4. Shot from 24 inches: only soot deposited on target

5. Shot from 25-36 inches: scattered specks of unburned/partially burned gunpowder grains

6. Shot from 3+ feet: usually no residue on target

Primer Residue on Hands - some primer residue goes back to shooter
- doesn't stay in hands very long
- dermal nitrate test - hot wax to pick up traces of nitrate, false positives (nitrate in urine and tobacco)
- primer residue test - no false positives, primer residue sticks to webbing of skin, closeby people may have it too, but the shooter has MORE, so AMOUNT matters

gunshot wounds:
1. temporal cavity - lasts millisecond - causes tremendous damage, tear tissue, destroy organs, can be as large as a cannonball, massive stretching caused by gases expanding as bullet passes thru body.
2. permanent cavity - crushing of tissue & bone caused by bullet itself & if only as large as bullet itself, will heal unless fatal

1. hollow point bullet: mushrooms on contact ² very large wound - fragments
Ie. civilian ammo
2. full metal jacket: no mushrooming or fragmentation
Ie. miliary ammo ² aim to disable, not kill


Questioned Documents (QD)
--------------------------------------
- Any object which contains signs, symbols or marks either visible, partially visible or invisible, which convey a meaning to someone
- Handwriting is unique if: reasonable amount of writing and executed freely and fluently

Graphology is NOT QD, Graphology is determining personality from handwriting, like fortune telling

Children - copies letter from books, parents, relatives
Adult – writing is a semi-conscious habit
Habits very hard to change or disguise

Class Characteristics
- from same copybook
- from same grade 1 teacher
- restricted to certain profession
- bubble font in north american teenage girls
- visually impaired?

individual characteristics
- due to diff perception
- diff in physical dexterity
- bad habbits
- from influencial people

Natural Variation
- Usual and normal deviations that occur within repeated specimens of a person’s handwriting e.g. signature – no two ever the same

Types of comparison samples:
1. collected samples - collected from everyday life
pros: undisguised, natural variation, contemporary?
cons: contain diff. text from questioned document, the real author?

2. requested samples - under warrant, watched by police
pros: knows real author, questioned document can be duplicated
cons: maybe disguised, nervous, not contemporary (at the same time as QD)

Obtaining Requested Samples
Dictate – rapidly, adjusting speed
Do not include punctuation
Must not see original
Seated comfortably (or as for QD)
Duplicate writing instrument and material
Repeat text, paragraph X3, signature X15
Remove specimen after written
Text must include all words, numbers of QD
Duplicate size of QD
Duplicate style of QD – e.g. letter
If printed, ensure upper and lower case both used

prove authorship: by witness, by someone familiar with suspect's writing

Opinions
Positive identification – sample and QD written by same person, strong probability, weak probability
Can neither identify nor eliminate
Negative – prove that someone else wrote document, writing quality is better than that which suspect is capable

Factors which affect handwriting comparisons
- pen, paper
- writing position
- condition of person writing, drunk?

characteristics - pen pressure, style, slanting

Disguise - how long can you keep a lie for?
Simple, not fluent, change is rarely consistent
Altered letter design, internal consistency disrupted
Never original – several basic types
Lapses back into own style, certain features never disguised

Class characteristics – make and model of machine, e.g. font size, design, pitch, shape
Innate characteristics – characteristic which is common to a group, but not all in group, e.g. dirt in a mould for manufacture of 10 printers
Individual characteristics – actual machine, e.g. damage to a key

Photocopiers
Document – make and model of machine
Actual machine identified by trash marks, picker marks, damage
New models – leave identifying mark
Number of copies since QD
Currency? Will shut down

Typewriters
Old ribbons – could read writing directly
Newer ribbons – moves up and down, can still be read

Alteration of Documents After Their Production
Torn paper/foil/plastic
Water soaked documents, charred documents
Altered or erased
Latent impressions

onto the land with no women

PDB

http://www.pdbwiki.org
http://www.prweb.com/releases/genomics/bioinformatics/prweb3865864.htm
http://www.openhelix.com/pdb

http://database.oxfordjournals.org/cgi/content/full/2010/0/baq009?etoc

Probability Overview

http://www.cs.utah.edu/~hal/courses/2009S_AI/Walkthrough/Probability/

Probability is the likelihood of an event occurring. This site discusses the following topics:

Random variables

Joint and conditional distributions

Invariance

Bayes’ rule

Independence

libgail problem, lost panels

http://tech.zhenhua.info/

It turned out that I need to use dpkg command
dpkg --remove –-depends libgtk2.0-0 libgtk2.0-0-common etc

Then use following command to fix the broken dependencies:
sudo apt-get install –f

http://ubuntuforums.org/archive/index.php/t-923414.html
sudo aptitude install nautilus

finally gave up and just re-enabled jaunty in /etc/apt/sources.lst and update and re-installed ...
bad idea to mix jaunty with intrepid ...

Definitions in AI

Rational Agent: For each possible percept sequence, a
rational agent should select an action that is expected
to maximize its performance measure, given the
evidence provided by the percept sequence and
whatever built-in knowledge the agent has.

Setting for intelligent design:
PEAS: Performance measure (safest), Environment (road), Actuators (horn, brake),
Sensors (speedometer, camera)

A problem is defined by four items:
1. initial state (Arad)
2. actions or successor functions (Arad->Zenrind)
3. goal test (Bucharest)
4. path cost (# of actions executed, nodes visited)

tree search - expanding states / nodes

completeness: does it always find a solution if one exists?
time complexity: number of nodes generated
space complexity: maximum number of nodes in memory
optimality: does it always find a least-cost solution?

uninformed search:
- breadth-first - fifo, go wide, time and space = O(b^d+1),
- depth-first - lifo, go deep, time=O(b^m), space = O(bm) (only keep one path in memory)
- depth limited
- iterative-deepening

Uninformed search - Summary
• Problem formulation usually requires abstracting away real-
world details to define a state space that can feasibly be
explored
•
• Variety of uninformed search strategies
•
• Iterative deepening search uses only linear space and not
much more time than other uninformed algorithms
•


A*
Evaluation function f(n) = g(n) + h(n)
g(n) = cost so far to reach n
h(n) = estimated cost from n to goal
f(n) = estimated total cost of path through n to
goal

Theorem: If h(n) is admissible, A* using TREE-
SEARCH is optimal

Local search:
- hill-climbing (climb mount everest in thick fog with amnesia)
- simulated annealing (allow bad moves but decrease their frequency)
- local beam search
- genetic algorithms

Informed-Search Summary
• Heuristic functions estimate costs of shortest paths
• Good heuristics can dramatically reduce search cost
• Greedy best-first search expands lowest h
– – incomplete and not always optimal
• A∗ search expands lowest g + h
– – complete and optimal
– – also optimally efficient (up to tie-breaks, for forward
search)
• Admissible heuristics can be derived from exact
solution of relaxed problems
• Admissible heuristics never overestimates true cost of the solution

constraint satisfaction problem (CSP)
- map-coloring
- nodes are variables
- edges are constraints

back-tracking search is uninformed depth-first search for CSPs

forward checking keeps track of valid legal moves, terminate when there's no more valid moves

arc consistency - An Arc X->Y is consistent if for every value x of X there is some value y in Y consistent with x

if a constraint graph has no loops, then CSP can be solved in O(nd^2) time
- linear in the number of variables.

2 general approaches to convert cyclic graphs to trees
1. Assign values to specific variables (Cycle Cutset method)
2. Construct a tree-decomposition of the graph
- Connected subproblems (subgraphs) form a tree structure

A CSP is k-consistent if for any set of k-1 variables and for any consistent
assignment to those variables, a consistent value can always be assigned to any kth
variable.

solve tree csps by
1. convert graph to tree, pick a parent, so any node only has one parent
2. apply backward pass (constraint propagation)
3. forward pass (assignment)

heuristics:
- Most constrained variable (highest number of edges): choose the variable (eg. SA map) with the fewest legal values
- Least constraining value - Given a variable, choose the least constraining value (eg. pick red)

Arc-consistency (AC) is a systematic procedure for constraining propagation

min-conflict heuristic for hill climbing in n-queens

CSP-Summary
• CSPs
– special kind of problem: states defined by values of a fixed set of variables, goal test
defined by constraints on variable values
• Backtracking=depth-first search with one variable assigned per node
• Heuristics
– Variable ordering and value selection heuristics help significantly
• Constraint propagation does additional work to constrain values and detect
inconsistencies
– Works effectively when combined with heuristics
• Iterative min-conflicts is often effective in practice.
• Graph structure of CSPs determines problem complexity
– e.g., tree structured CSPs can be solved in linear time.

Minimax strategy
Find the optimal strategy for MAX assuming an infallible MIN
opponent
– Need to compute this all the down the tree

alpha-beta pruning
Depth first search – only considers nodes along a single path
at any time

The Horizon Effect
– sometimes there’s a major “effect” (such as a piece being
captured) which is just “below” the depth to which the tree has
been expanded

Expected Minimax = sum_over_chance_nodes(prob(x)+minimax(x))

Minimax - Summary
• Game playing can be effectively modeled as a search problem
• Game trees represent alternate computer/opponent moves
• Evaluation functions estimate the quality of a given board
configuration for the Max player.
• Minimax is a procedure which chooses moves by assuming that
the opponent will always choose the move which is best for
them
• Alpha-Beta is a procedure which can prune large parts of the
search tree and allow search to go deeper
• For many well-known games, computer algorithms based on
heuristic search match or out-perform human world experts.

logic:
m is a model of a sentence alpha if alpha is true in m

alpha |= b, alpha entails sentence b if and only if b is true in all worlds where
alpha is true.

Model-checking enumerates all possible worlds

If an algorithm only derives entailed sentences it is called
sound or truth preserving.

Completeness : the algorithm can derive any sentence that is
entailed.

truth table connectives


Two sentences are logically equivalent iff they are true in same models: α ≡ ß
iff α╞ β and β╞ α

Logical equivalence


A sentence is valid if it is true in all models,
e.g., True, A V ~A, A => A, (A ^ (A => B)) => B
(tautologies)
Validity is connected to inference via the Deduction Theorem:
KB ╞ α if and only if (KB => α) is valid
A sentence is satisfiable if it is true in some model
e.g., A V B, C
(determining satisfiability of sentences is NP-complete)
A sentence is unsatisfiable if it is false in all models
e.g., A ^ ~A
Satisfiability is connected to inference via the following (aka proof by contradiction):
KB ╞ α if and only if (KB ^ ~α) is unsatisfiable

Normal form = conjunctions of disjunctions

Horn clause = A clause with at most 1 positive literal. e.g. A V ~B V ~C = B ^ C => A

1 positive literal: definite clause

Forward chaining (sound and complete for Horn KB)
Idea: fire any rule whose premises are satisfied in the KB,
– add its conclusion to the KB, until query is found

Backward chaining (BC)
check if q is known already, or
prove by BC all premises of some rule concluding q
Hence BC maintains a stack of sub-goals that need to
be proved to get to q.
- Like FC, is linear and is also sound and complete (for Horn KB)

• FC is data-driven, automatic, unconscious processing,
– e.g., object recognition, routine decisions
• May do lots of work that is irrelevant to the goal
• BC is goal-driven, appropriate for problem-solving,
– e.g., Where are my keys? How do I get into a PhD program?
• Complexity of BC can be much less than linear in size of KB

Model Checking (Satisfiability - SAT problems)
Two families of efficient algorithms:
• Complete backtracking search algorithms: DPLL algorithm
• Incomplete local search algorithms
– WalkSAT algorithm (random flipping, only good if we know a solution exists)

Logic - Summary
• Logical agents apply inference to a knowledge base to
derive new information and make decisions
• Basic concepts of logic:
– syntax: formal structure of sentences
– semantics: truth of sentences wrt models
– entailment: necessary truth of one sentence given another
– inference: deriving sentences from other sentences
– soundness: derivations produce only entailed sentences
– completeness: derivations can produce all entailed
sentences
• Resolution is complete for propositional logic
• Forward, backward chaining are linear-time, complete
for Horn clauses
• Propositional logic lacks expressive power

• Propositional logic assumes the world contains facts,
• First-order logic (like natural language) assumes the world
contains
– Objects: people, houses, numbers, colors, baseball games, wars,
...
– Relations (returns True or False - Facts / atomic sentence): Brother(Richard, John)
– Functions (returns another object): LeftLegOf(John), Sqrt(3)
– Variables x, y, a, b,...
– Connectives , , , ,
– Equality =
– Quantifiers ,

Note: Functions do not state facts and form no sentence:
– Brother(Pete) refers to John (his brother) and is neither true nor false.

for all Vx King(x) => Person(x)
there exists Ei Integer(i) ^ GreaterThan(i,0)

Squares are breezy near a pit:
– Diagnostic rule --infer cause (pit / cavity) from effect (breeze / toothache), given a breeze, what's the chances of a pit?
s Breezy(s) r Adjacent(r,s) Pit(r)
– Causal rule---infer effect from cause (model based reasoning), given a pit, what's the chances of a breeze?
r Pit(r) [ s Adjacent(r,s) Breezy(s)]

First-order logic Summary:
– Much more expressive than propositional logic
– Allows objects and relations as semantic primitives
– Universal and existential quantifiers
– syntax: constants, functions, predicates, equality, quantifiers
–
• Knowledge engineering using FOL
– Capturing domain knowledge in logical form

propositionalization (convert FOL to propositional form, problem is generates lots of irrelevant sentences, fix is to use unification):
- Universal instantiation (UI) : Subst({v/g}, α) means the result of substituting ground term g for variable v in sentence α, produces a whole set of instantiated sentences
e.g King(John) Greedy(John) Evil(John), {x/John}

- Existential instantiation (EI) : For any sentence α, variable v, and constant symbol k (that
does not appear elsewhere in the knowledge base):
E.g., x Crown(x) OnHead(x,John) yields:
Crown(C1) OnHead(C1,John)
where C1 is a new constant symbol, called a Skolem constant

Idea for doing inference in FOL:
– propositionalize KB and query
– apply resolution-based inference
– return result

Unification (better way to propositionalize than instantiation):
• Recall: Subst(θ, p) = result of substituting θ into sentence p
• Unify algorithm: takes 2 sentences p and q and returns a
unifier θ if one exists
Unify(p,q) = θ where Subst(θ, p) = Subst(θ, q)
• There is a single most general unifier (MGU) that is unique up
to renaming of variables.
eg. Knows(John,x) and Knows(y,z) MGU θ = { y/John, x/z }

Knows(x,OJ) and Knows(John,x) {fail}!!! because x can't be both OJ and John, fix is to standardize variable (rename x to z so Knows(z,OJ) and Knows(Jonh,x) )


Modus ponens: p, p->q
-------
q

Generalized Modus Ponens (GMP) (slide 13)
p1', p2', ... , pn', ( p1 ^ p2 ^ ... ^ pn -> q)
------------------------------------------------
Subst(θ,q)
where we can unify pi', and pi for all i
Example:
p1' is King(John) p1 is King(x)
p2' is Greedy(y) p2 is Greedy(x)
θ is {x/John,y/John} q is Evil(x)
Subst(θ,q) is Evil(John)


GMP is complete for a KB consisting of definite clauses ( a ^ b -> c = not a V not b or c , 1 positive literal)

• Forward-chaining
– Uses GMP to add new atomic sentences
– Useful for systems that make inferences as information streams in
– Requires KB to be in form of first-order definite clauses
• Backward-chaining
– Works backwards from a query to try to construct a proof
– Can suffer from repeated states and incompleteness
– Useful for query-driven inference
• Resolution-based inference (FOL)
– Refutation-complete for general KB
• Can be used to confirm or refute a sentence p (but not to
generate all entailed sentences)
– Requires FOL KB to be reduced to CNF
– Uses generalized version of propositional inference rule

Prolog
• Closed-world assumption ("negation as failure")
– e.g., given alive(X) :- not dead(X).
– alive(joe) succeeds if dead(joe) fails // so assume joe is alive if he's not dead

Program = set of clauses = head :- literal1, ... literaln.
criminal(X) :- american(X), weapon(Y), sells(X,Y,Z), hostile(Z).

Appending two lists to produce a third:
append([],Y,Y).
append([X|L],Y,[X|Z]) :- append(L,Y,Z).

append(input1, input2, output)

• append(A,B,[1,2]) ? // what input value combinations for A and B will give [1,2] output?
• answers: A=[] B=[1,2]
A=[1] B=[2]
A=[1,2] B=[]

Resolution:
(A V B V C), ~A
-----------------
(B V C)


(A V B V C), ( ~A V D V E)
--------------------------
(B V C V D V E)


FOL inference Summary
• Inference in FOL
– Simple approach: reduce all sentences to PL and apply
propositional inference techniques
– Generally inefficient
• FOL inference techniques
– Unification
– Generalized Modus Ponens
• Forward-chaining: complete with definite clauses
– Resolution-based inference
• Refutation-complete

Eliminate existential quantifiers (Skolemize):
∃x P(x) becomes P(K) (EE)
K is some new constant (Skolem constant)

∀x∃y P(x,y) becomes ∀x P(x,F(x))
F() must be a new function (Skolem function)

Probabilistic assertions summarize effects of
– Laziness - failure to enumerate exceptions, qualifications, etc
– Ignorance - lack of relevant facts, initial conditions, etc.

Default or nonmonotonic logic:
– Assume my car does not have a flat tire

Utility theory is used to represent and infer preferences

Decision theory = probability theory + utility theory

The fundamental idea of decision theory is that an agent is rational if
and only if it chooses the action that yields that highest expected
utility, averaged over all the possible outcomes of the action.

Atomic event: A complete specification of the
state of the world about which the agent is
uncertain
E.g., if the world consists of only two Boolean variables
Cavity and Toothache, then there are 4 distinct
atomic events:
Cavity = false Toothache = false
Cavity = false Toothache = true
Cavity = true Toothache = false
Cavity = true Toothache = true

Axioms of probability
• For any propositions A, B
• 0 ≤ P(A) ≤ 1
– P(true) = 1 and P(false) = 0
– P(A V B) = P(A) + P(B) - P(A ^ B)

In the study of probability, given two random variables X and Y defined on the same probability space, the joint distribution for X and Y defines the probability of events defined in terms of both X and Y

unconditional / prior probability P(A) : probability A will appear in the absence of any other information, eg. P(Cavity) = 0.1

conditional / posterior probability P(A|B) : probability of A changes with respect to new information B eg. P(Cavity|Toothache) = 0.8
P(A|B) = P(A^B)/P(B) can be rewritten as product rule: P(A^B)=P(A|B)P(B)=P(B|A)P(A)

if A and B are independent, then P(A|B)=P(A) so P(A^B)=P(A|B)P(B)=P(A)P(B)

Finetti (2 player game, player1 has subjective probability 'a' on the occurrence of 'b'):
- subjective beliefs should respect the axioms
- If the beliefs are contradictory, then the agent will fail in its environment
in the long run!
- if bet $0.4 for b and P(b)=0.4, if b happens (0.4 chance), we take other player's money who bet $0.6 so 0.4*0.6 and if b doesn't happen (0.6 chance), we loose $0.4 so -0.6*$0.4 so
0.4*$0.6 - 0.6*$0.4 = 0
- so we either pay how much we bet, or win the bet of the other player

Cavity, Toothache, Catch Joint probability distribution

P(toothache) = 0.108 + 0.012 + 0.016 + 0.064 = 0.2

P(toothache or cavity) = P(toothache) + P(cavity) - P(toothache and cavity) = (0.108 + 0.012 + 0.016 + 0.064) + (0.108 + 0.012 + 0.072 + 0.008) - (0.108 + 0.012) = 0.28

P( ~cavity | toothache) = P( ~cavity ^ toothache) / P(toothache)
= (0.016+0.064) / (0.108 + 0.012 + 0.016 + 0.064)
= 0.08 / 0.2 = 0.4

P( cavity | toothache) = P( cavity ^ toothache) / P(toothache)
= (0.108+0.012) / (0.108 + 0.012 + 0.016 + 0.064)
= 0.120 / 0.2 = 0.6

General idea: compute distribution (X) on query variable (Y) by fixing evidence
variables (E) and summing (adding all possible combinations) over hidden variables (H), so H = X - Y - E

• Obvious problems (fix is independence!, divide and conquer):
1. Worst-case time complexity O(dn) where d is the largest arity
2. Space complexity O(dn) to store the joint distribution
3. How to find the numbers for O(dn) entries?

Normalization / Relative likelihood:
let alpha = 1 / P(toothache), hidden variable catch, then
P( cavity | toothache) = alpha * P( cavity ^ toothache) = alpha * P(cavity,toothache,catch) + P(cavity,toothache,~catch) = alpha * (0.108+0.012) = alpha * 0.12
**note, when alpha is involved, need to normalize the final value and to do this, need to find out both P(a) and P(~a) so alpha = 1/(p(a) + p(~a))

Similarly,
P( ~cavity | toothache) = alpha * P( ~cavity ^ toothache) = alpha * P(~cavity,toothache,catch) + P(~cavity,toothache,~catch) = alpha * (0.016+0.064) = alpha * 0.08

To get real probability values,
P(cavity|toothache) + P(~cavity|toothache) = 1
alpha * (0.12+0.08) = 1, alpha = 5, so
P(cavity|toothache) = 5*0.12 = 0.6
P(cavity|toothache) = 5*0.08 = 0.4

Independence: P(Toothache, Catch, Cavity, Weather) = P(Toothache, Catch, Cavity) P(Weather)
So instead of 2*2*2*4=32 entries, it's broken down to 2*2*2+4=12 entries

Catch is conditionally independent of Toothache given Cavity:
• P(Catch | Toothache,Cavity) = P(Catch | Cavity)
So the probe catching it isn't affected by whether or not we have a toothache,

Baye's rule for multiple evidence (toothache, catch) (assuming catch is conditionally independent of toothache given cavity) is:
P(toothache, Cavity, Catch) = alpha * P(toothache|cavity) * P(catch|cavity) * P(cavity)

Baye's Rule: P(Y|X)P(X) = P(X|Y)P(Y) or P(Y|X) = P(X|Y)P(Y) * alpha where alpha = 1/P(X)
useful as P(cause|effect) = P(pit|breeze) = P(cavity|toothache) = P(effect|cause)*P(cause)/P(effect)

This is an example of a naïve naive Bayes model:
• P(Cause,Effect1, ... ,Effectn) = P(Cause) πiP(Effecti|Cause)
or P(Effect1, ... ,Effectn|Cause) = πiP(Effecti|Cause)

Uncertainty / Probability - Summary
• Probability is a rigorous formalism for uncertain
knowledge
• Joint probability distribution specifies probability
of every atomic event
• Queries can be answered by summing over
atomic events
• For nontrivial domains, we must find a way to
reduce the joint size
• Independence and conditional independence
provide the tools


Definition of Bayesian networks
- Representing a joint distribution by a directed graph (nodes=random variables, edges = direct dependence)
- key is to look at parents: p(X1, X2,....XN) = π p(Xi | parents(Xi ) ) = CPT conditional probability tables
- Can yield an efficient factored representation for a joint distribution


chain rule for probabilities
P(a, b, c, … z) = P(a | b, c, …. z) P(b, c, … z) = P(a | b, c, …. z) P(b | c,.. z) P(c| .. z)..P(z)

law of total probability
P(a) = Sum_over(b) P(a | b) P(b) where B is any random variable

Conditional Independence:
P(a | b, c) = P(a | c) tells us that learning about b, given that we already know c, provides no change in our probability for a,
i.e., b contains no information about a beyond what c provides

Constructing a Bayesian Network: (E=earthquake, B=burglary, A=alarm, J=John calls, M=Mary calls)
Step 1. Order the variables in terms of causality e.g., {E, B} -> {A} -> {J, M}
Step 2. p(X1, X2,....XN) = π p(Xi | parents(Xi ) ) so P(J, M, A, E, B) = P(J | A) P(M | A) P(A | E, B) P(E) P(B)

Marginal Independence:
p(A,B,C) = p(A) p(B) p(C)

A Tale of 3 graphs:
1. {c->b and c->a} Conditionally independent effects:, a & b are not independent
p(A,B,C) = p(A|C)p(B|C)p(C) -- a & b are not independent
e.g. disease c causes symtoms b and a
but p(A,B|C) = p(A|C)p(B|C) -- a & b are dependent when conditional on C, p(C) cancels out

2. {a->c->b} {earthquake->alarm->john calls} markov dependence,
p(A,B,C) = P(b|c)p(c|a)p(a) -- a & b are not independent
but p(A,B|C) =

3. {a->c, b->c} (earthquake->alarm, burglar->alarm} Independent Causes: a & b are independent
p(A,B,C) = p(C|A,B)p(A)p(B)

If we have a Bayesian network, with a maximum of k parents for any node, then we need O(n*2^k) probabilities (instead of O(2^n))

Markov blanket
A node is conditionally independent
of all other nodes in the network
given its Markov blanket (in gray) (blanket includes parents, children and spouses)

Polytree: there is at most one undirected path between any two nodes. Like Alarm. (opposite of multiply connected graph - like a diamond-shape graph). Time and space complexity in such graphs is linear in n

Approximate inference, Weather example:


Give that exact inference is intractable in large networks. It is essential to consider approximate inference models

i. Discrete sampling / prior sampling method - generation of samples from a known distribution, sample each variable in a topological order, conditioning each variable using the values of it parent, problem - need large sample size??
eg. like flipping a coin 1000 times and count the number of heads
eg. suppose the order is [Cloudy, Sprinkler, Rain, Wet Grass] and we suppose we get [True,False,True,True] in the sampling, then the probability is P(Cloudy)*P(~Sprinkler|Cloudy=T)*P(Rain|Cloudy=T)*P(WetGrass|Sprinkler=F,Rain=T) = 0.5 * 0.9 * 0.8 * 0.9 = 0.324., so we should expect to see [T,F,T,T] around 324 times out of 1000.

ii. Rejection sampling method - Is a general method for producing samples from a hard to sample distribution., first it generates samples from P(X|e), rejects those that don't match the evidence, count X in the remaining samples, problem is it rejects too many samples, hard to sample rare events (red sky at night)
eg. find P(rain|sprinkler=true) using 100 samples, suppose we get 73 sprinkler = false, then we throw those away and only look at the remaining 27 sprinkler = true, say 8/27 rain = true, then the probabiliy is 8/27=0.296, the true answer is 0.3.

iii. Likelihood weighting - avoids the problem with rejection sampling by generating events only consistent with e in P(X|e), fixes the values of evidence variables and generates samples for non-evidence variables conditioned on the parents, each event is weighted by the likelihood that it happens with the evidence, 2 parts: evidence - use weight, non-evidence - use same, combine the two by multiplying all of them
http://www.cs.utah.edu/~hal/courses/2009S_AI/Walkthrough/LikelihoodWeighting/


iv. MCMC (Markov-chain Monte-Carlo) algorithms (Gibbs sampling)
- Unlike other samplings which generate events from scratch, MCMC makes a random change to the preceding event.
- At each step a value is generated for one of the non evidence variables condition on its markov blanket.
- Assume that calculating p(x|markovblanket(x)) is easy
- randomly wanders around the state space, changing non-evidence variables one at a time while keeping evidence variables fixed
e.g. p(rain|sprinkler=true, wetgrass=true), so evidence variables are sprinkler and wetgrass
0. initialize non-evidence variables cloud and rain, say true and false, so [true, true, false, true]
1. sample cloudy non-evidence variable conditioned on its markov blanket (children, spouse and parent, so sprinkler and rain). so p(cloudy|sprinkler=true, rain=false), suppose cloudy is false then new state is [false, true, false, true]
2. rain is sampled, then p(rain|cloudy=false, wetgrass=true, sprinkler=true), suppose rain=true, then new state is [false, true, true, true]

this step is repeated, then say we found 20 states where rain is true, and 60 states where rain is false, then p(rain|sprinkler=true, wetgrass=true)=<0.25,0.75>