Pattern Hunter faster and more sensitive homology search

  • Slides: 19
Download presentation
Pattern. Hunter: faster and more sensitive homology search Bin Ma, John Tromp and Ming

Pattern. Hunter: faster and more sensitive homology search Bin Ma, John Tromp and Ming Li Bioinformatics Vol. 18 no. 3 2002 Oxford University Press

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References 1

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References 2

BLAST § Basic Local Alignment Search Tool § Steps: 1. Seeding 2. Extension 3.

BLAST § Basic Local Alignment Search Tool § Steps: 1. Seeding 2. Extension 3. Evaluation 3

BLAST: Seeding PAM BLOSUM T > 10 PGQ (18) -> PAQ (12) PGQ (18)

BLAST: Seeding PAM BLOSUM T > 10 PGQ (18) -> PAQ (12) PGQ (18) -> PGA (11) … 4

BLAST: Extension 1 -hit extension method 2 -hit extension method Cluster near and nonoverlapped

BLAST: Extension 1 -hit extension method 2 -hit extension method Cluster near and nonoverlapped hits (< A). Gapped Extension Smith-Waterman algorithm variation. 5

BLAST: Evaluation 6

BLAST: Evaluation 6

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References 7

Pattern. Hunter § Steps 1. Seeding 2. Extension 3. ¿Evaluation? 8

Pattern. Hunter § Steps 1. Seeding 2. Extension 3. ¿Evaluation? 8

Pattern. Hunter: Seeding Optimal Spaced Seeds for Hidden Markov Models, with Application to Homologous

Pattern. Hunter: Seeding Optimal Spaced Seeds for Hidden Markov Models, with Application to Homologous Coding Regions. Brejová, B. , Brown, D. , Vinar T. 9

Pattern. Hunter: Extension 1 -hit extension method If 1 -hit fails 2 -hit extension

Pattern. Hunter: Extension 1 -hit extension method If 1 -hit fails 2 -hit extension method Cluster near and nonoverlapped hits (< A). Gapped Extension Red-Black Tree algorithm variation. 10

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References 11

1 -hit performance comparison 12

1 -hit performance comparison 12

2 -hit performance comparison 13

2 -hit performance comparison 13

Performance comparison 14

Performance comparison 14

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References 15

Conclusions § Pattern. Hunter offers better results than BLAST, in terms of computation and

Conclusions § Pattern. Hunter offers better results than BLAST, in terms of computation and search sensitivity. § Unlike BLAST, Pattern. Hunter runs on all platforms that support Java. § BLAST algorithm, and its internal behavior, is widely known in bioinformatics. § Unlike Pattern. Hunter, BLAST is licensed for public domain. § Homology search algorithms growing usage. 16

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References

Index § Introduction • BLAST • Pattern. Hunter § Comparison § Conclusion § References 17

References § http: //en. wikipedia. org/wiki/BLAST § http: //bioinformatics. oxfordjournals. org/cgi/reprint/18/3/44 0. pdf §

References § http: //en. wikipedia. org/wiki/BLAST § http: //bioinformatics. oxfordjournals. org/cgi/reprint/18/3/44 0. pdf § http: //citeseerx. ist. psu. edu/viewdoc/download? doi=10. 1. 1. 1. 2393&rep=rep 1&type=pdf § http: //books. google. es/books? hl=es&lr=&id=MRdpjt. Db. Uc QC&oi=fnd&pg=PR 11&dq=blast+oreilly&ots=1 ROkssxu. F f&sig=Lp 2 r. NR 5 Zs. OIe. I 7 Bbt 8 Pu. ZTu. Qp 24 18