Genome annotation Assessment in Drosophila melanogaster

1
Genome annotation Assessment in Drosophila
melanogaster

• By Lauren Gonzalez
• Jeff Hu
• Cathy Zhou

2
Purpose

• The goal of this genome annotation assessment
  project (GASP) was to refine the accuracy of gene
  annotation technologies so that we could use it
  to our benefit to facilitate genome sequencing
  and annotation.

3
Identifying STDs

• The researchers created three standard sets of
  DNA by which the experimental results were
  compared to.
• std1
• std2
• std3

4
Standard Sets

• std1 Accurate position of base pairs, but might
  not include every gene
• std2 Original 80 cDNA sequences
• Std3 Has most genes, but exact placements might
  not be known

5
6 classes of gene annotation

• 1. Gene finding
• 2. Promoter prediction
• 3. Repeat finders
• 4. Protein homology annotation
• 5. EST/cDNA alignment
• 6. Gene function

6
A comparison of testing results
7
Understanding results

• True positive
• True negative
• False positive
• False negative

8
Sensitivity vs. Specificity

• Sensitivity ratio of true positives to all real
  positives (ratio of true hits)
• TP / (TP FN)
• Specificity ratio of true positives to all found
  positives (ratio of real hits)
• TP / (TP FP)

9
Algorithmic Problem Testing

• We will investigate the ability of a java code to
  identify repeat sequences in different ORFs in a
  strand of DNA. Our threshold for a repeat
  sequence will be different for each repeat
  sequence length
• Length Threshold
• 1 base pair 4 in a row
• 2 base pairs 8 in a row
• 3 base pairs 12 in a row
• 4 base pairs 16 in a row

10
APT cont.

• We will then take the number of repeat sequences
  (and their respective lengths) that are on or
  above threshold, and analyze which repeat
  sequence has the highest occurrence. We will run
  this java program on many different strands of
  DNA, and look at which strands have higher repeat
  rates, and how reading in different ORFs change
  our results.