Complex disease gene mapping Part I

hey =D

i'm not purposely to post something like this.
i was just... being so sien of reading off my notes as usual...
and actually the genetic test will be on 26 june...
gonna start preparing immunology exam from tomorrow on until 22 june...
i think reading + typing + thinking will enhance my memory...
HWAHAHAHA
so i decided to type while i'm looking through the notes...
u don't have to read this post tho. it's kinda boring.

alright. stop crapping.
here we go.

ok. why do we need to map the disease genes? how?

There are so many medically important human diseases developing nowadays, most of them are complex and multifactorial.
We have to find out the genetic susceptibility factors in order to diagnose the disease in the early stage, predict the prognosis and thus the best treatments can be designed to for the patient.
For complex diseases, the research includes:
family/twins/adoption studies to test if there's significant genetic component to the disease,
linkage analysis (affected sib pair analysis) to map the susceptibility loci,
refine the candidate gene region by population associations,
identify the DNA variants conferring susceptibility, and
define the biochemical action of the susceptibility alleles.

Linkage analysis looks for alleles shared by affected individual, without using models, within families or in a whole population, in two ways:
Identity By Descent- to prove alleles are copies of the same parental allele
Identity By State- look at identical alleles but cannot prove common ancestry
but for some rare alleles, its unlikely to say that they are from independent origins, so IBS implies IBD... how we define IBD?
->we use multi-locus and multi-allele haplotypes microsatellite markers because they are more efficient than two-allele markers.

What is haplotype?
It's the haploid genotype, 'the genotype of a single chromosome', a collection of associated alleles on a chromosome segment. Everyone has two haplotypes for each of a particular genome section.

Affected sib pair analysis (ASP)...
by random segregation, sib pairs have 1/4 chance to share 0 parental haplotype, 1/2 chance to share 1 parental haplotype, and 1/4 chance to share 2 parental haplotypes.
hmm...this is a bit hard to explain from the chart -.-...
example:
Daddy: AB; Mummy: CD
they have the 1st child, Adam: AC
Mummy is now pregnant again,then we are to calculate the chance for the 2nd child, Eve, to share the alleles with Adam...
Possible outcomes: AC,AD,BC,BD
If Eve is AC, which means she and Adam share two parental haplotypes;
if Eve is AD, they share 'A', which is one parental haplotype;
if Eve is BC, they share 'C', which is one parental haplotype; and
if Eve is BD, they do not have the same parental haplotypes.
This explains: sib pairs have 1/4 chance to share 0 parental haplotype, 1/2 chance to share 1 parental haplotype, and 1/4 chance to share 2 parental haplotypes.
=D

So,
if both sibs are affected by a genetic disease, they are likely to share the locus that carries the disease...
therefore,
if they are both carrying a mutant allele at the locus,
if the disease is recessive, they share two parental haplotypes;
if the disease is dominant, they share at least one parental haplotype.

Advantages:
it is genotyped for many markers to look for sharing of loci above 1:2:1 ratio.
The more sib pairs, the more powerful the analysis =.="
multipoint analysis is preferred because it's more efficient. (IBD sharing)
Simple and powerful analysis for non-Mendelian disease susceptibility genes.
Disadvantages:
The candidate region of genes is is usually large.
It cannot be reduced by testing closer and closer markers. =(
Not all sib pairs will share the locus if the susceptibility factor is neither necessary nor sufficient.

Association studies and Linkage disequilibrium (LD)...
Association is the statistical statement about the co-occurence of alleles or phenotypes.
Example:
If people with disease A have significantly more or less of allele B than the predicted from know frequencies, then disease A is associated with allele B.
The causes can be contribution of direct causation, selective advantage, population stratification (disease and 'associated allele' happen to shared by a particular subdivision), LD and......
Transmission Disequilibrium Test is to avoid the stratification by using family based controls and trios (affected proband and two parents). (number of times the allele is transmitted to offspring compared to other alleles?)
Linkage= genetic difference between loci; association= relationship between alleles and phenotypes. Linkage creates association within families, not among unrelated people.

Argh. I'm not sure if this is important---International HapMap Project?!
this is siao. the theory is... funny yet relevant... well. not really funny... okok. i'll just briefly go through it. hahaha...

The project samples come from 270 people from 4 populations. 90 Yoruba individuals (30 trios) from Ibadan, Nigeria, so called YRI? -.-" 90 of Europeans descent from Utah... CEU??? 45 Han Chinese from Beijing, CHB... @.@ and 90 Tokyo Japanese (JPT)...
Their DNA samples are converted into cell lines...
In all 270 people,
Phase I: 1 million SNPs genotyped-> at least one common SNP/5kb acroos the whole genome!
Phase II: another 4.6 million SNPs genotyped
the project contributes 6 million new SNPs!

And the results they've got...
The genetic evidence shows that all humans today, yea, you, me, they...all of us!... are descended from modern ancestors in Africa around 150,000 years ago...
Most of the current human variation present in the ancestral population...as we migrated out of Africa, we took part of the genetic variation, so-called bottleneck. Why is it? Because they found that the haplotypes seen outside Africa are likely to be the subsets of the haplotypes inside Africa, and the haplotypes in non-African populations happen to be longer due to the less recombination. -.-"
Frequency of haplotypes is different in different populations due to random chance, natural selection... and also, mutations create new haplotypes... Newer haplotypes maybe found only in a single population...
Marker LD systemic studies: chromosomes contain a series of islands of long-range LD sharply separated from each other. LD may extend 50kb within an island, but there's no markers between islands although they are close. These island boundaries are hotspots for recombination. From this, we can find haplotypes at each set of markers and test the association with any diseases. Then, we don't need to test so many markers... We only compare the haplotypes from affected individuals to the individuals without disease. If any particular haplotypes happen to be more frequent in the affected people, we can locate the disease associated gene within or near the haplotype. Therefore, we need to look at haplotype frequencies in multiple populations first, then only we choose the molecular markers for the particular haplotype identification.

You see. I think the idea of this project comes from seriously bored intelligent people. cia ba bo su zo.

ok. now it's the important part. SNPs. Single Nucleotide Polymorphisms. The variation of single nucleotide due to the common mutations- substitution, deletion or watever!
SNPs. the most abundant form of DNA variation occurs approximately one in 200bp. It's more abundant although it is less informative than micro and mini satellites markers. SNPs are not spread evenly, they are clustered, more susceptible to mutation eh?! The cluster of SNPs can be called haplotype also hor, if it's inherited as a unit.
Tag SNPs representing a particular haplotype. "relatively small set of variants that capture most of the common patterns of variation in the genome".--->then we don't need cover the whole genome loh.
We can use them as an indirect association approaching tools to find any candidate gene in genome, or any region from family-based linkage analysis, or scan whole genome for disease risk factors.

Limitations:
1. common SNPs with large spacing used-> lose of important information if the disease causing variants are rare.
2. we don't know how well the SNPs represent other genetic variants like repetitive elements or indels.
3. identification of genetic risk factors not necessarily brings improvements in health.
well. those hardworking geneticists are still finding the way out! i mentally support you guys. XD

ethical issues......
sien nia. can i just stop here. i'll continue with part II tomorrow.
will start my fave immunology tomorrow.
won't be this suffering! @#$#@$%#$%#@$

my legs are cramped. my face is dry. my stomach is aching. my eyes are tired (watched too much CSI NY older season...). and i am off to sleep now!!! yay!
but so shit. i didn't do exactly what i have planned for today T.T
CSI ruined me T.T
hwahahaha. anyway, i deserve it. sigh. Eddie Cahill is cute. he looks like Josh Hartnett @.@
But then the last episode i watched just now is about the bombing of one building...
Eddie Cahill as det. Flack got severe injured @.@
that scene was so geli!
his chest was broken until the blood vessels are exposed!!!
not out of his body la. luckily Mac was there...

shit. hav to stop here. =.="
i think this is the longest post i've ever written!
full of craps!!!
ciaoz. -.-
good night...