HMMSCAN output
Hello,
I am trying to understand the output of HMMSCAN. I went trough the documentation and I understand the lines returned by the program.
In the following example, how can I deduce in which state is the model at each column? For each position in the seq, I would like to extract the probability distribution over the 20 possible AA given by the HMM model, however I don't know whether the model is in a match or insert state.
Query: d1ux8a_ [L=119]
Description: a.1.1.1 (A:) Protozoan/bacterial hemoglobin {Bacillus subtilis [TaxId: 1423]}
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
------- ------ ----- ------- ------ ----- ---- -- -------- -----------
2.9e-32 111.5 0.0 3.2e-32 111.3 0.0 1.0 1 Bac_globin Bacterial-like globin
0.00063 19.8 0.0 0.00098 19.2 0.0 1.2 1 Protoglobin Protoglobin
Domain annotation for each model (and alignments):
>> Bac_globin Bacterial-like globin
# score bias c-Evalue i-Evalue hmmfrom hmm to alifrom ali to envfrom env to acc
--- ------ ----- --------- --------- ------- ------- ------- ------- ------- ------- ----
1 ! 111.3 0.0 3.3e-36 3.2e-32 3 120 .. 4 117 .. 2 118 .. 0.94
Alignments for each domain:
== domain 1 score: 111.3 bits; conditional E-value: 3.3e-36
HHHHTHHHHHHHHHHHHHHHHHT-CCCCCCGGGTTSSHHHHHHHHHHHHHHHTTSSSHHHHHHSS-.HHHHHTTS--BHHHHHHHHHHHHHHHHHCT-- CS
Bac_globin 3 ydaiGgeatveelvdrfYervdedeeealseifsgtdlegsreklveFlvqalgGPelYverrGhPrlrkrHapfeitaaerdawlkhlaeAleetgvs 101
y+aiG + + +lvd fYerv++ + l++if ++dl++ k+ +Fl+q+lgGP+lY+e +GhP+lr+rH+pf+it++ +dawl ++++A++++g +
d1ux8a_ 4 YEAIGE-ELLSQLVDTFYERVASHPL--LKPIF-PSDLTETARKQKQFLTQYLGGPPLYTEEHGHPMLRARHLPFPITNERADAWLSCMKDAMDHVGLE 98
999986.99*****************..88887.789999999******************************************************** PP
HHHHHHHHHHHHHHHHHHC CS
Bac_globin 102 sedreeklellealahaaq 120
e re + +le +a ++
d1ux8a_ 99 GEIREFLFGRLELTARHMV 117
**99999999999888875 PP
• 915 views
•
link
0 answers
No answers yet.
Log in to answer this question.