[Bioperl-l] BUG
fangl
fangl at big.ac.cn
Mon Apr 15 08:13:05 UTC 2013
Hi,
I find a bug when I use Bio::SearchIO to parse Hmmscan.
The result of my Hmmscan has 2 hits against 1 query ,but the module of Bio::SearchIO can only parse 1 hit in the output file.
eg:
hmmscan result:
# hmmscan :: search sequence(s) against a profile database
# HMMER 3.0 (March 2010); http://hmmer.org/
# Copyright (C) 2010 Howard Hughes Medical Institute.
# Freely distributed under the GNU General Public License (GPLv3).
# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
# query sequence file: pro_2.fa
# target HMM database: /leofs/biodenovo/fangl/data/Pfam/Pfam-A.hmm
# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Query: PEG03688 [L=200]
Scores for complete sequence (score includes all domains):
--- full sequence --- --- best 1 domain --- -#dom-
E-value score bias E-value score bias exp N Model Description
------- ------ ----- ------- ------ ----- ---- -- -------- -----------
0.00038 20.7 0.1 0.00054 20.2 0.1 1.3 1 SnoaL_2 SnoaL-like domain
------ inclusion threshold ------
0.068 12.9 0.0 0.43 10.3 0.0 2.0 1 FAD_binding_8 FAD-binding domain
Domain annotation for each model (and alignments):
>> SnoaL_2 SnoaL-like domain
# score bias c-Evalue i-Evalue hmmfrom hmm to alifrom ali to envfrom env to acc
--- ------ ----- --------- --------- ------- ------- ------- ------- ------- ------- ----
1 ! 20.2 0.1 7.9e-08 0.00054 11 86 .. 59 157 .. 46 172 .. 0.72
Alignments for each domain:
== domain 1 score: 20.2 bits; conditional E-value: 7.9e-08
T-HHHHHHTEEEEEEEE............CT.S--..E---HHHHHHHTTHHHCEECEEEEEEEEESSTTEEEEEEEEE-E........ESBS--EEEE CS
SnoaL_2 11 gdldalaallapdvvwe............dp.fge..lrGrealraffrallaafpdlrfevedviadgdrvvvrwtvtgt........ipptgrgvtv 86
d l +l++ d+++ +p +g+ + G++a+ + f ++ + f+++++++d+ v++++ +t+t i p ++ +v
PEG03688 59 PDYNLLKELVTYDCTYIsltfdnptlhgiMPwAGThtHVGPQAFIDIFTRVGLYWDRGPFSIDHIFGDDGNVTAWGSFTATsrtlgktvISPWAARARV 157
3666888999999999877777555545545544488******************************************98555555554444444444 PP
>> FAD_binding_8 FAD-binding domain
# score bias c-Evalue i-Evalue hmmfrom hmm to alifrom ali to envfrom env to acc
--- ------ ----- --------- --------- ------- ------- ------- ------- ------- ------- ----
1 ? 10.3 0.0 6.3e-05 0.43 19 63 .. 77 128 .. 63 197 .. 0.69
Alignments for each domain:
== domain 1 score: 10.3 bits; conditional E-value: 6.3e-05
FAD_binding_8 19 lklkkpkks.......lkykpGqyvfini.pslsklflqsHPFtiasapeddk 63
l++ p+ ++ + G +fi i ++ l+++ PF+i dd
PEG03688 77 LTFDNPTLHgimpwagTHTHVGPQAFIDIfTRVG-LYWDRGPFSIDHIFGDDG 128
555555444444444468899999*****66666.99********88777664 PP
Internal pipeline statistics summary:
-------------------------------------
Query sequence(s): 1 (200 residues)
Target model(s): 13672 (2396357 nodes)
Passed MSV filter: 244 (0.0178467); expected 273.4 (0.02)
Passed bias filter: 219 (0.0160181); expected 273.4 (0.02)
Passed Vit filter: 14 (0.00102399); expected 13.7 (0.001)
Passed Fwd filter: 2 (0.000146284); expected 0.1 (1e-05)
Initial search space (Z): 13672 [actual number of targets]
Domain search space (domZ): 2 [number of targets reported over threshold]
# CPU time: 0.15u 0.13s 00:00:00.28 Elapsed: 00:00:06.72
# Mc/sec: 71.32
//
my parse result:
query len hit len query_start query_end query_match_len hit_start hit_end hit_match_len evalue cov
PEG03688 200 SnoaL_2 0 59 157 99 11 86 76 7.9e-08 0.49
pleast to solve this problem!
Thanks!
Best wishes!
fangl
More information about the Bioperl-l
mailing list