A2.2

Question 1 Single best answer Level 1 1 point

What is a pileup at a given reference position?

  1. The column of bases that all reads covering one position contribute
  2. The count of reads covering each position, plotted along the genome
  3. The row of bases one read contributes along its whole length
  4. The consensus sequence assembled from reads overlapping a region

Question 2 Single best answer Level 1 1 point

A read shows a base that differs from the reference at one position. What would make that difference a biological mutation rather than a sequencing error?

  1. The differing base was introduced while the library was amplified by PCR
  2. The differing base was misreported by the sequencer during the run
  3. The differing base sits at the end of the read, where alignment is poor
  4. The differing base is present in the DNA molecules of the sampled cells

Question 3 Single best answer Level 2 1 point

A substitution and a two-base deletion occur in the same region of a sample. Every read is placed at the correct locus, but in the reads covering the deletion the aligner has matched the bases straight through without opening a gap. A pileup reports the substitution cleanly in one column, and reports the deletion as a run of mismatches continuing past it. What is it about the deletion that produces a run rather than a single column?

  1. It removes two bases, so two columns are simply left empty in those reads
  2. It shifts every later base out of the column holding its reference position
  3. It lowers the alignment score, so the aligner marks the whole region mismatched
  4. It changes the bases on either side of it, creating real mismatches nearby

Question 4 Single best answer Level 2 1 point

A caller weighs homozygous reference against heterozygous at a position covered by 30 reads, 8 of which carry the alternate base. Why can the read counts alone not settle which genotype is correct?

  1. The heterozygote predicts 15 alternate reads, so 8 rules it out and leaves homozygous reference
  2. Homozygous alternate also fits 8 of 30, so a third genotype is still in contention
  3. Both genotypes can produce 8 alternate reads, one by uneven sampling and one by error
  4. Homozygous reference predicts no alternate reads, so 8 rules it out and leaves the heterozygote

Question 5 Single best answer Level 2 1 point

A deduplication tool flags reads sharing identical mapping coordinates as duplicates of one original molecule. Two independent DNA molecules happen to break at the same two points, so their reads map to identical coordinates. What could we give the tool that the coordinates cannot?

  1. A count of how many reads share each coordinate, which the tool then thresholds
  2. A quality score per read, so the tool discards the less reliable of the two
  3. A tag added before amplification, so reads from two molecules stay distinguishable
  4. A second alignment pass that moves the reads to their true distinct coordinates

Question 6 Single best answer Level 2 1 point

A candidate heterozygous variant is supported by 20 alternate reads, all of which are PCR duplicates of a single original molecule. The caller treats each of those 20 as a separate observation. What has gone wrong?

  1. Twenty copies of one observation are being counted as twenty observations
  2. Twenty duplicates map to slightly different positions, disturbing the alignment
  3. Twenty duplicates carry lower base quality than the reads they were copied from
  4. Twenty duplicates genuinely deepen the coverage, so the site is well supported

Question 7 Single best answer Level 2 1 point

A caller reports a genotype at a site with GQ 20, where \(GQ = -10\log_{10}(P(\text{error}))\). What does that number assert?

  1. About a 1 in 100 chance the variant at this site is not real
  2. About a 1 in 20 chance the genotype assigned here is wrong
  3. About a 1 in 100 chance the genotype assigned here is wrong
  4. That about 1 in 100 of the reads at this site disagree with the call

Question 8 Single best answer Level 3 1 point

At a candidate heterozygous site, the reads carrying the alternate base all have low base quality, while the reads carrying the reference base have high base quality. The number of reads on each side is unchanged. What happens to the likelihood of the heterozygous genotype, and to the call?

  1. It is unchanged, because the same number of reads still carry the alternate base
  2. It falls, and the call moves toward homozygous reference
  3. It falls, but the call is unchanged because the prior still favors the heterozygote
  4. It rises, because uncertain bases leave more room for a variant to be present

Question 9 Single best answer Level 3 1 point

Site 1 is a single C-to-T change where every read covering it aligns cleanly end to end. Site 2 is a four-base deletion inside a stretch where those same four bases repeat several times, so the deletion can be written at more than one position. Which site needs a haplotype-aware caller rather than a pileup, and why?

  1. Site 2, because the deletion has no single column the whole signal sits in
  2. Site 1, because a pileup column cannot hold the context a substitution needs
  3. Neither, because more reads at each site would settle both without reassembly
  4. Both, because the aligner’s grid is too unreliable to call either site on
Back to top