I have been attempting to generate a PSSM using a local installation of blast that I have on a linux box. the command I am using is psiblast -db ~/blasdb/nr -in_msa alignedbinding -inclusion_ethresh 0.001 -out_ascii_pssm pssm.txt.
Running this as written it continually states "BLAST engine error: PSSM has positive average score". When you try a different matrix from the default (PAM70 for example) it claims it returns no hits from my database which I know to be false as I tested the database using the same sequences and all returned hits using blastp that were orders of magnitude below the evalue threshold specified..
MSA is relatively large (7130 sequences), formatted in Clustal format in UGENE on my local windows box and (according to the paper) the data has set been used successfully by another research group so there should (in theory at least) be no issues so I think it more likely that I have screwed up somehow but cannot spot it..
Can anyone assist please as this is maddening and I need that PSSM.
just an update. If you add -ignore_msa_master to that command it runs for a little while and produces the start of pssm.txt but then quickly reverts to complaining about positive average score.