[HOME] [DESCRIPTION] [HELP] [NEWS] [CONTACT] [Geno3D]
July 30, 2024: NPS@ updated (see NEWS).
[Abstract] [NPS@ help] [Original server]
Paste a multiple protein alignment in Clustal W format below : help
Restrict profile HMM HMMBUILD to multiple sequence alignment boundaries : from column to
Profile HMM HMMBUILD options : Alternative model construction strategies (only --fast supported): - Symbol fraction in each MSA column (--symfrac, 0<x<1): - Fraction of MSA columns to tag sequence as fragment (--fragthresh, L <= x*alen, 0<x<1): Alternative relative sequence weighting strategies: Henikoff position-based weights (--wpb) Gerstein/Sonnhammer/Chothia tree weights (--wgsc) Henikoff simple filter weights (--wblosum) don't do any relative weighting; set all to 1 (--wnone) (--wgiven not supported) - Henikoff simple filter weighting strategy (--wblosum): identity cutoff (--wid, 0<x<1) Alternative effective sequence number strategies: adjust eff seq # to achieve relative entropy target (--eent) eff seq # is # of single linkage clusters (--eclust) no effective seq # weighting: just use nseq (--enone) set eff seq # for all models to <x> below (--eset) - eff seq # to achieve relative entropy target (-eent): minimum relative entropy/position (--ere) (bits/position), sigma param (--esigma): - eff seq # is # of single linkage clusters (--eclust): fractional identity cutoff (--eid, 0<x<1): - eff seq # for all models (--eset): Alternative prior strategies: default don't use any prior; parameters are frequencies (--pnone) use a Laplace +1 prior (--plaplace) Control of E-value calibration: - MSV Gumbel mu fit: length of sequences (--EmL) , number of sequences (--EmN) - Viterbi Gumbel mu fit: length of sequences (--EvL) , number of sequences (--EvN) - Forward exp tail tau fit : length of sequences (--EfL) , number of sequences (--EfN) , tail mass (--Eft,0<x<1) Other options: - Seed (--seed): - Inserts length (--maxinsertlen)