This is the BLAST(Basic Local Alignment Search Tool) graph. How can I calculate its time complexity?
Asked
Active
Viewed 1,229 times
2
-
**Picture is attached above** – LiaNalani Dec 23 '19 at 11:15
-
Probably O(Queries*Sequences)? – Ctx Dec 23 '19 at 11:16
-
1@Someprogrammerdude Hm, in my opinion this picture isn't particularly suitable for a text version... – Ctx Dec 23 '19 at 11:21
1 Answers
3
BLAST algorithm steps:
Creating a list of words of length W of the query sequence.
Search for W words in the database.
Elongation of hit sequences, ie those found, and assignment of a score. These sequences will be given by a local alignment.
Taking the so called HSP (High-scoring Segment Pair).
The First step take time O(n), where n is the number of element in the sequence.
The Second step is another loop on the words created in the first step ( upperbound n) so time O(n)
The Third step assign an hit score. This could be done only letter by letter so the time complexity is O(n*m) where m is the number of letter in the query word. As upper bound we can say that the time is O(n^2)
The fourth step is a simple cycle over the score obtained, so ever O(n).
In conclusion the algorithm could be assimilated to O(n^2), in the case m << n, we could say O(n*m). the lower bound is O(n).
Zig Razor
- 3,381
- 2
- 15
- 35