I discovered this site after being put in touch with both the authors of this site by a friend in the lab, and following the link in the e-mail signature. I was looking for help with Perl programming, and the BioPerl? extensions (desperately trying to automate a bioinformatic task which would have taken weeks otherwise).
After arriving I had a strange sense of deja vu, until I realised I had seen Requiem on this site a year or two ago.
Anyhow. Perl programming is going well; what a mad language! However, if anyone can advise me on the best nucleotide alignment program to use for a series of non-flush sequences which overlap in some way, then it would be appreciated. I have tried ClustalW?, T-COFFEE and some internal BioPerl? objects, all without success as of yet.
Have you tried BLAST yet? You can blast two sequences against each other (well... query sequence against the second sequence as though the second sequence is the BLAST database). I find it rather more reliable than ClustalW?, although you can only do two sequences at once. Here: http://www.ncbi.nlm.nih.gov/blast/bl2seq/bl2.html - SunKitten
Presumably you can use BioPerl? to submit BLAST queries. If you're mad enough to want to write your own, Google: approximate string matching would be the place to start, but bear in mind such things tend to require *lots* of CPU time - it's the sort of thing you want to run on a bunch of networked machines. There's probably little point reinventing the wheel - if you really want to run something locally, you can always [download] a copy of BLAST. - MoonShadow
Actually, what I am doing is using BLAST both locally and remotely. I have a local database of contigs assembled from ESTs. I BLAST my short sequences against this database, and then want to take the full sequences of the best hits and assemble them together into a larger sequence. This can then be remotely BLASTXed against the non-redundant database at the NCBI to find homologies. The reason I am doing this is because the original short sequences return no homology from BLASTX, so with a longer sequence we have a better chance of getting some results. BLAST can't really be used for the alignments since it is only a local alignment program, and won't give the free ends of the aligned sequences. I have managed to get the best success with ClustalW? for the aligments though, in terms of actual returned results (25 out of 125 sequences have come back with some homologies with low E-values). Anyhow, BioPerl? seems pretty cool - it handles remote and local BLASTing very well, and has an object model for interpreting reports. My next task is automatically generating Gene Ontology annotations for 7000 sequences. Could be fun! Thanks for your help! - Andrew
And finally: I am not going to get addicted to this site.
Thank you. Very quick! What's with all these question marks? It's as if the machine knows the subjects I should next write pages on! Is it artificially intelligent, or something less clever? --Andrew
Words with capital letters in the middle are automatically assumed to be page titles. They become links if the page already exists, and obtain question marks otherwise. Check out WhatIsWiki, the "What are all these weird blue question marks?" section. ..oh, and RecentChanges is how people spot when people post :) - MoonShadow
Welcome to ToothyWiki. You don't have to be mad... ah, you get the picture. --FR
Is it cheeky to put a link to your own webpage (http://www.srcf.ucam.org/~ajm226) here? I dropped off Google's index, and I need to find a quick and easy way of getting back on!
![[Home]](/img/text_logo.jpg){kind=logo}
Andrew