On Wed, 6 Oct 2004, Kevin Karplus wrote: > >An HMM is a good way to model a domain family, but using HMMer without >calibration is likely to result in poor models. You'll get better OK. The thing is, sould you ever buid a model to find higly similar sequences, or should you just use blast to do that? >models using the SAM package with the provided target2k script for >builing the multiple alignment, followed by the w0.5 script to do >appropriate thinning and sequence weighting for building a >superfamily-level model. If you need to end up with a HMMer model, >you're still better off building the model with SAM and using >sam2hmmer. > >If you don't pay careful attention to thinning and/or sequence >weighting, then models built from the usual redundant alignments are >terrible. > > >------------------------------------------------------------ >Kevin Karplus karplus at soe.ucsc.edu http://www.soe.ucsc.edu/~karplus >Professor of Biomolecular Engineering, University of California, Santa Cruz >Undergraduate and Graduate Director, Bioinformatics >Senior member, IEEE Board of Directors, ISCB (starting Jan 2005) >life member (LAB, Adventure Cycling, American Youth Hostels) >Effective Cycling Instructor #218-ck (lapsed) >Affiliations for identification only. >