Sequence alignment and reading a search result · seed 1 · A4, ink-friendly. The answer key prints on its own page for grown-ups.

Scoring alignments and judging hits

Science · Genetics & Evolution · ages 20-22
Name ______________________   Date ____________
  1. A BLAST hit has E = 2. What does that value mean?

    • Two chance alignments at least this strong are expected
    • The hit has exactly a two-percent probability of being false
    • Two biological functions have been confirmed for the hit
  2. For candidate alignments using the same scoring scheme and constraints, which rule selects the best score?

    • Choose the longest alignment regardless of penalties
    • Choose the alignment with the highest total score
    • Choose the alignment with the most gap characters
  3. Why can a short query have a large E-value even when its aligned letters match perfectly?

    • Perfect identity makes database size irrelevant
    • A short sequence can also occur by chance
    • Every perfect short match contains a sequencing error
  4. In one BLAST search, hit A has E = 0.000001 and hit B has E = 2. Which statement follows from these values?

    • Hit B has stronger support because its E-value is larger
    • Hit A's exact function is now experimentally established
    • Hit A has stronger evidence against a chance match
  5. Use +2 per match, -1 per mismatch and -2 per gap character. An alignment has 5 matches, 2 mismatches and 1 gap character. What is its total score?

    Answer: ______________

  6. With +2 per match, -1 per mismatch and -2 per gap character, an alignment with 2 matches, 1 mismatch and no gaps scores 3.

    Circle one:   True   False

  7. Use +2 per match, -1 per mismatch and -2 per gap character. For 4 matches, 2 mismatches and 1 gap, a friend writes 8 + 2 = 10 and stops. What repairs the calculation?

    • Add every penalty as positive and leave the gap unscored
    • Use 8 - 2 and omit the penalty for the gap character
    • Use 8 - 2 - 2 = 4, subtracting both penalties
  8. A long query has a tiny E-value for an alignment covering only a short region. A student assigns the hit's full protein function to the query. What is still needed?

    • Evidence that the matched region supports that function
    • A rule that any tiny E-value proves the full protein function
    • A larger E-value before any biological evidence is considered
LightMySky · lightmysky.comW1-mt_jNbEMFqgZ5-s1

Answer key

For grown-ups. Fold this page away before handing over the rest.

Scoring alignments and judging hits W1-mt_jNbEMFqgZ5-s1

  1. Two chance alignments at least this strong are expected · The E-value is an expected count of chance alignments meeting or exceeding the observed score in the search space.
  2. Choose the alignment with the highest total score · The scheme defines contributions from all columns. The highest total is preferred under the same constraints.
  3. A short sequence can also occur by chance · Perfect identity over a few letters can occur by chance. Identity and statistical support describe different things.
  4. Hit A has stronger evidence against a chance match · The smaller expected chance count supports hit A more strongly under the model. It does not establish a function.
  5. 6 · The contributions are 10, -2 and -2. Their sum is 6.
  6. True · Two matches contribute 4, one mismatch contributes -1 and gaps contribute 0. The total is 3.
  7. Use 8 - 2 - 2 = 4, subtracting both penalties · The calculation added the mismatch cost and omitted the gap cost. Both contributions are negative.
  8. Evidence that the matched region supports that function · Strong statistical similarity in one region does not establish the full protein's function. Check alignment coverage and functional evidence.
Worksheet · LightMySky