[Bioperl-l] RE:Segmentation fault

stephan rosecker stephan.rosecker at ish.de
Mon Apr 12 14:01:29 EDT 2004


 > I really have no idea what you are doing....

:-)
I got an protein-profile-file build with hmmer. (nst.hmms).
So i use hmmpfam with a flatfile database "uniprot_sprot.dat" to find
proteins compatible with my profile.
(hmmpfam -A0 -E1e-30 nst.hmms uniprot_sprot.dat > hmmerresultfile.out)
So each result contains a hit or more is what i need.



This for example is good: (hmmerresultfile.out)

//

Query sequence: 128U_DROME
Accession:      P32234
Description:    GTP-binding protein 128UP.

Scores for sequence family classification (score includes all domains):
Model    Description                                    Score    E-value  N
-------- -----------                                    -----    ------- ---
Drg                                                     639.4   1.6e-191   1

Parsed for domains:
Model    Domain  seq-f seq-t    hmm-f hmm-t      score  E-value
-------- ------- ----- -----    ----- -----      -----  -------
Drg        1/1       4   367 ..     1   375 []   639.4 1.6e-191

Alignments of top-scoring domains:
Drg: domain 1 of 1, from 4 to 367: score 639.4, E = 1.6e-191
                    *->GIeEkIkaIEeEiarTqKNKATEyHLGlLKaKLAkLReQLlEpskgg
                        I+EkI aIE E+arTqKNKAT  HLGlLKa  AkLR +L   s +g
   128U_DROME     4    -ILEKISAIESEMARTQKNKATSAHLGLLKANVAKLRRELI--SPKG 47

                    gGGG...kGFeVeKSGdarvaLIGfPSVGKStLLakLTntkSEvAdYaFT
                    gGGG+++ GFeV+K Gdarv ++GfPSVGKStLL+ L +  SEvA+Y+FT
   128U_DROME    48 GGGGtgeAGFEVAKTGDARVGFVGFPSVGKSTLLSNLAGVYSEVAAYEFT 97

                    TLTcvPGvLeYqgakIQllDlPGIIeGAssGKGRGrqVlAvarsADLvLm
                    TLT vPG ++Y+gakIQllDlPGIIeGA  GKGRGrqV+Avar+  L+ m
   128U_DROME    98 TLTTVPGCIKYKGAKIQLLDLPGIIEGAKDGKGRGRQVIAVARTCNLIFM 147

                    vLDatkleehrdvLekELEnVGIRLnkrkGPniyfKkKetGGvsvngTgp
                    vLD  k+ +h  +Le ELE  GIRLnk++ Pniy+K K  GG++ n+ +p
   128U_DROME   148 VLDCLKPLGHKKLLEHELEGFGIRLNKKP-PNIYYKRKDKGGINLNSMVP 196

                    PLThidEdtvrrILheYRIhNAeVliReDqvTlDdFIDVVnegNRvYipc
                       ++d d+v++IL+eY+IhNA++  R D +T Dd IDV+ egNR Yipc
   128U_DROME   197 -QSELDTDLVKTILSEYKIHNADITLRYD-ATSDDLIDVI-EGNRIYIPC 243

                    LyVyNKIDlvgieevdrlarkrlEiGpntIVvScnmglnLdelaerlwek
                    +y +NKID+++iee+d + +      p  + +S+    n d+l+e +we
   128U_DROME   244 IYLLNKIDQISIEELDVIYKI-----PHCVPISAHHHWNFDDLLELMWEY 288

                    LglvRVYtKkqGkePDFDPLvepLilRrGsT.VeDvChhiHrDLvrrFKY
                    L l R+YtK++G+ PD+     p +l +  T++eD+C+++Hr  +++FKY
   128U_DROME   289 LRLQRIYTKPKGQLPDYN---SPVVLHNERTsIEDFCNKLHRSIAKEFKY 335

                    AlVWGkSaKhppQrVGleHvleDeDvvqivvK<-*
                    AlVWG+S+Kh pQ+VG+eHvl DeDvvqiv+K
   128U_DROME   336 ALVWGSSVKHQPQKVGIEHVLNDEDVVQIVKK    367

//


This for example is useless: (hmmerresultfile.out)

//
Query sequence: 10KD_VIGUN
Accession:      P18646
Description:    10 kDa protein precursor (Clone PSAS10).

Scores for sequence family classification (score includes all domains):
Model    Description                                    Score    E-value  N
-------- -----------                                    -----    ------- ---
         [no hits above thresholds]

Parsed for domains:
Model    Domain  seq-f seq-t    hmm-f hmm-t      score  E-value
-------- ------- ----- -----    ----- -----      -----  -------
         [no hits above thresholds]

Alignments of top-scoring domains:
         [no hits above thresholds]
//





best regards,
stephan







 >
 >
 > If you want to initialize a SearchIO parser to read in HMMER output
 > results you do
 >
 > my $in = Bio::SearchIO(-file => 'hmmerresultfile.out',
 >                        -format => 'hmmer');
 >
 > # an loop through each result (a single result per query sequence)
 > while( my $result = $in->next_result ) {
 >   # print the query name
 >   # for hmmpfam this will be each sequence in the input set
 >   # for hmmsearch this will be the name of the HMM file you
 >   # used to search
 >   print $result->query_name, "\n";
 >
 >   # now print all the hits (for HMMPFAM this will be the Pfam domains
 >   # for hmmsearch this will be the sequences in the db that had good
 >   # alignments
 >   while ( my $hit = $result->next_hit ) {
 >     # print the name, indented with a tab
 >     print "\t", $hit->name, "\n";
 >   }
 > }




More information about the Bioperl-l mailing list