More genomics: Basic assignment

The answers to the basic part are due Tue, March 21.

Genome alignments: Blanchette et al

1. Why did the authors use synthetic data in their comparison of TBA against other multialigment programs?
2. In table 1 of the paper, which report running times for different programs and datasets, MULTIZ has "Not Applicable" (NA) reported for the dataset simulated for nine mammals. Why is that?
3. In Figure 4A, it appears that all tested programs did very well on the Human-Baboon comparison. Why is that?
4. Consider three DNA sequences with pairwise alignments as follows (for q0 vs q1, and q0 vs q2):

   >q0
   ATAGTGACAGTGCCCTAAG---CAAACACACGCTT-----GGAGCAGGTCGTCATGCGTG
   >q1
   AAGTGGATAC-----GGAGCTGCAAACACACACTTGCTCATCAGCAGGATGTCCACCCTG
   

   and

   >q0
   ATAGTGACAGTGCCCTAAGCAAACACACGCTT-----GGAGCAGGTCGTCATGCGTG
   >q2
   -----GATAGTGCCTTTCACAAAGACACACTTGCTCAGGAGCAGGATGTCCACCCTC
   

   According to the star alignment method (look it up if you did not understand it), how will the three-way alignment with q0 as center turn out?

GWAS

5. Give actual examples of
   1. selection bias
   2. misclassification bias
6. Is there reason to believe that sequencing-based GWAS are more powerful than SNP-based studies? Motivate.
7. What are the main reasons for why some diseases' genetic associations are very hard to find?
8. In what way can multi-marker analysis help when a causative SNP is actually not among those SNPs that we genotype?