nodalpoint.org - software - Comments

You're welcome

Neil — Thu, 04 Oct 2007 03:44:26 -0400

You're welcome and apologies for omitting "-c 0.9" in my first comment. I'm sure you worked it out.

Thanks !

pansapiens — Wed, 03 Oct 2007 22:37:17 -0400

Thank you Pawel and Neil.

CD-HIT is exactly what I want, and it works great. The algorithm appears to take some smart short-cuts and reduces calculation time from would have taken many hours with my half written brute-force all-against-all pairwise alignment script down to seconds. If CD-HIT is good enough for UniProt, it's good enough for me :)

Blastclust looks like another good option (it was even already installed on my machine ... right under my nose), but is orders of magnitude slower, so I'm sticking with CD-HIT for my quick-n-dirty tasks.

Lucky I didn't go very far coding something to do this myself (although thinking about the problem was enlightening). I've got to remember to listen to my internal "stop-immediately-you-are-reinventing-the-wheel" alarm more often.

I second cd-hit

Neil — Wed, 03 Oct 2007 07:10:33 -0400

CD-HIT works for me. Simple as "cd-hit -i file -o file90 -c 0.9 -n 5".

Sorry about your comments Pawel, they went to spam for some reason.

cd-hit, blastclust

pawel — Wed, 03 Oct 2007 06:43:08 -0400

I'm not sure if my previous comment was saved... Anyway, I will point again to cd-hit and blastclust. However, the first does not make any alignments, and allows some redundancy in the set anyway.

MAFFT

nuin — Thu, 31 May 2007 13:04:32 -0400

Yep, MAFFT is fast and has excellent alignment quality. My last year's paper showed that it is the best for prtein alignment, even for distant sequences.

quicktree

jason — Wed, 30 May 2007 19:07:49 -0400

I'd use MAFFT for aligning something this big.
quicktree was designed for building trees from large datasets like this (Pfam families).

Phylip

nuin — Mon, 28 May 2007 12:07:55 -0400

Apart from CD-HIT which is a very good software, I would try using Phylip. The sequences are not large but the actual number of sequences is the problem. Try using a Neighbor Joining approach in Phylip, it won't be blazing fast but it will do the job, eventually.

I used Phylip to build a NJ tree of a set of 20000 protein sequences and it took me around 3-4 weeks to get it done on a 3GHz Xeon machine.

cd-hit

Neil — Sat, 26 May 2007 09:48:06 -0400

For clustering, CD-HIT is excellent. Very fast, handles many sequences. Used to create the non-redundant datasets in UniProt and at the PDB.

Phylogeny - I've never gone much beyond Clustal and Phylip, both of which would take hours on an average machine with any more than a few thousand sequences. I've heard good things about MrBayes - which is MPI-enabled, so could run on a cluster if you have access to one.

Huge!

Animesh — Sat, 26 May 2007 01:25:17 -0400

Alf that is a huge requirement! Are you trying to make tree of life with 1000 base upstream of some house-keeping genes?
I think MUSCLE [ http://www.drive5.com/muscle/ ] can come to rescue, but you would need a good machine for sure. It uses log-expectation as profile function which is faster and accurate as well [ http://www.biomedcentral.com/1471-2105/5/113/table/T2 ]. General algorithm is http://nar.oxfordjournals.org/content/vol32/issue5/images/large/gkh340f2... .
More details in the paper http://nar.oxfordjournals.org/cgi/content/full/32/5/1792 .

______________________"The Answer Lies in Genome"______________________
http://computationalbiologynews.blogspot.com/

Taverna

mart1nus — Wed, 11 Apr 2007 16:35:06 -0400

I don't get it. The diagram is too complex? Of course it is complex. But everything is pointing towards a workflow (SOA) world. Why does bioinformatics deny this turn? Oracle, SAP, IBM and so on all rewrite all their applications so they are able to use BPEL as graphical buil environment for integration processess. So stick in the 90s and keep using scripts, but within a few years we all will use workflows in a SOA world. I really believe this will happen, IBM believes in it so does the whole IT world. Let's see who's right.

Re: new blast interface

dalloliogm — Tue, 03 Apr 2007 10:35:12 -0400

Finally!!
I've never liked the actual interface ;)
In particular, I hated the pop-up window to show the results and I think the interface of the actual blast is not too clear.
The new feature to keep track of the recent results sounds cool.
Thanks ;)

---

WHO AM I?
Your name is , seeker.

http://genome.imim.es/~giovanni

A hack (not too kludgy)

pjw — Wed, 28 Mar 2007 09:02:33 -0400

Designate one of the multiple output as a representative file (say foo.psl), touch the other files at the end of the commands, and have the rule: foo.log foo.whatever ...: foo.psl

Doesn't change a thing

Antonio Piccolboni — Fri, 23 Mar 2007 00:22:26 -0400

What you wrote unfortunately is equivalent to what I wrote. It's like a shorthand notation for two separate rules, I should have explained it right away. It is going to execute the command twice. What we need is a way to express that a command has multiple outputs, and rules with multiple targets don't accomplish that.

makeovers

Ian Holmes — Wed, 21 Mar 2007 12:26:55 -0400

We certainly would want the replacement to look almost (if not exactly) like make. It should definitely be as simple as make -- that's probably the primary design goal: that you can cut and paste from the command line to a pipeline description file, with minimal extra typing. I'm fully aware of the dangers of excessive re-engineering...

Thanks for the other links... we'll certainly look into them, and post discussion summaries on biowiki.

http://biowiki.org/IanHolmes

Why not scripting?

Duncan — Wed, 21 Mar 2007 04:20:40 -0400

Scripting is great, a powerful tool that lets you achieve world peace in three lines of PERL/Python/Ruby. But what if people don't want to hack scripts? According to Grady Booch, the history of software engineering is one of increasing levels of abstraction, which is where Taverna and workflows are trying to go. Admittedly, we're not quite there yet, sometimes the abstractions leak, and as Stew says, you end up hacking BeanShell, thats not really a problem with Taverna or workflows, its an inherent problem in bioinformatics data, a flat-file legacy nightmare, that means we'll be forced to hack scripting languages for a long time to come, whether we like it or not.