SNPs

For the life of me I can not prevent myself from being distracted (when I really really need to study for Human Genetics).
So let's skip the review session and write a blog about it!

This particular lecture is focused on SNPs (it's fun to say snips!) They are important in identity tests, and locating genes that contribute to disease.
So... Single Nucleotide Polymorphism - because instead of having an A like me, you have like a G in the same spot on your DNA.
On average there's 1 different nucleotide out of every 1,000.
Microsatellites on the other hand are more thinly spaced along the genome (also a polymorphic marker, but instead of a single nucleotide difference, I would have a ATATAT while you would have ATAT).

Of the small amount of SNPs located on exons, (coding SNPs), they can be synonymous or nonsynonymous depending on whether the single base change in that 3 letter code results in the same or a different amino acid. If a different amino acid is coded, it is likely to change the function of the protein and cause disease.

To identify (genotype) SNP, you start with PCR
1. amplify DNA with SNP
2. separate the 2 alleles somehow, probably into major and minor alleles depending on the allele's frequency of occurrence.

There are several methods to chose from when identifying SNP.

The first choice presented in the lecture was primer extension. For this I feel like you need to know exactly where the SNP is already (from genome browser websites), so you can make a PCR product that's right up to, or almost at the SNP site.
You put in nucleotides that will add on to the primer through the action of enzymes, including one dideoxy nucleotide (like ddATP) so that after that one binds, no other nucleotides can attach to the DNA chain anymore. If there is a difference between the 2 alleles, the product will be different sizes as the one with a T opposite the ddATP would terminate early, while the other allele that didn't have a T would keep going until it runs into a T later on. The products can then be separated by gel, automated sequencer, or mass spectroscopy.

Second choice is to use sequencing chips. Here your PCR product is fluorescently labeled. You then try to hybridize it to a bunch of oligonucleotides (25 nucleotide sequences that code for DNA of the PCR product) attached to a glass slide in an array. In each column, the oligonucleotides would be shifted over by one nucleotide so that it slowly moves down the sequence of PCR product. In each row, the middle nucleotide of the oligo would be varying (T,G,C or A).
Then here's what I was confused about, but figured out last night.
So instead of having 1 reference sequence that you are basing the oligos off of, you are using 2 different reference sequences, the difference being the nucleotide you believe in different in the SNP. If you don't do it this way, and the PCR product don't match your reference sequence, then it won't hybridize to the oligos (with the exception of that one column where the difference is in middle of the oligo so that it will bind to the row with the appropriate base variation. (picture to come?)

A third way is kinda ghetto. You find SNPs by seeing if it adds or takes away a restriction enzyme recognition site. If one allele does and the other doesn't, when you digest the PCR product, they would be different sizes. But how many SNPs just happens to be at a restriction site. And also, incomplete digestion would lead to false results.

Linkage disequilibrium is a term describing the linkage between haplotypes. (haplotypes being alleles along one chromosome, while genotype is 2 chromosomes).
Linkage disequlibrium of SNPs occurs when there are no recombinations between the SNPs. This can be due to new mutations, or when SNPs are located in recombination cold spots.

Linkage disequilibrium is what you can use to identify disease genes. By picking a SNP every 20 to 40 kb (a SNP per haplotype block where no recombination occurs), much of the DNA can be covered to do an association test. The SNPs that are common in diseased individuals show an association, indicating the the gene affecting the disease is located close to that SNP. (examples of disease with identified genes: asthma, heart disease, schizophrenia, macular degeneration, etc.).

Take the neuregulin gene and schizophrenia for example.
Neuregulin 1 gene is a synaptic protein in the central nervous system.
Schizophrenia is a non-Mendelian (caused by multiple genes) common disease that affects 1% of the population.
Stefansson et al. 2002, 2003 associated the gene to disease in Icelandic and Scottish populations, and the knockout mouse (without the gene) has "similar" behavior and drug responses to humans with schizophrenia.
Yang et al. 2003 then went on to look at the gene in Chinese population.
He compared the SNPs of patients and their parents at the gene position.
The parents that are heterozygous would have genotype 1-2 for the SNP allele. If their child with the disease inherit allele 1 or 2 with equal frequency (50/50) then there is no association of that SNP with disease gene. If there is a difference in the inheritance of one SNP over another, then there is a positive association of the gene near the SNP to disease.

But of course, association does not = causation. Next up is to find the mutation that may cause the neuregulin gene to not work normally in the schizophrenic patients.

gee really... really???

After staying up till 5am today writing a grant that's all about the negative health consequences of being up at night, what do I see today from Gizmodo's working-the-night-shift-is-hazardous-to-your-health page?



A picture of Yuki and Kyo (they sure do look exactly like my cats, even in mannorism), who did indeed stay up all night with me being unhealthy. But wait... cats are nocturnal, so it's ok. But wait, a single night time light exposure results in days and days and days of circadian disruption. According to my grant, even if we go right back to a good sleep schedule (and supposedly good light dark cycles), its still weeks and weeks before our rhythm goes back to normal. And yes, my grant is very convincing with preliminary data and all, so you better watch out when I do the actual experiment!

Sleep Ed

Remember those Sex Ed/health classes in school? In my case, 'specialists' came to our school with models, and full length videos of the birthing process in glorious detail...
Now imagine that instead of being educated awkwardly about sex, the class was about sleep instead. For one of the gym class electives I was able to chose 'relaxation' which was nap time in high school. That was great! However sleep education is a little different...



In the issue of Sleep that came out yesterday, the article Evaluation of a School-Based Intervention for Adolescent Sleep Problem, described how they went to schools, told sleep deprived kids that not getting enough sleep is bad, and that they should sleep better. Not surprisingly, being told that you should sleep better did not make students sleep more in the long run.

I remember the motto in middle school was "sleep is for the weak". In order to finish your work and have some fun in the day, there is no way you can sleep for more than 6 hours. Once we got to high school, it became "sleep is a luxury". Since I wanted to do well in class, and practice flute and piano and be class officer for multiple clubs, and do sports sometimes too there's no way I can go to bed before 3am or 4am everyday just to wake up at 7am to go to school again. (I couldn't even sleep in on the weekends...T_T) Then college came around, and well, sleep just didn't exist anymore if we just wanted to pass and do nothing else.
So now that I am a grad student researching circadian biology, even knowing all the horribly negative effects of circadian disruption and sleep deprivation would not have changed my sleep habits in the past.

Sadly, effects of aging is finally catching up to me. I can no longer stay up for 4 days straight and feel perfectly fine. But that's ok, as a grad student it is now possible to go home after 10pm, chill for a few hours and be in bed to sleep for a whole 7 hours. But wait, the journal tells me to sleep 9hrs a day? Damn, I fails at doing what I preach.