Ok, here's where I, as the "microbial" person on this blog start speaking another language: most microbiologists, especially those who conduct large-scale environmental sampling for novel lineages of bacteria, archaea, and microbial eukaryotes, still use (gasp!) small subunit ribosomal RNA sequences for identifying organisms and analyzing communities. The reason is simple - primers to amplify these genes are almost completely universal so can be counted on to pick out even rare ("unculturable" almost goes without saying) microbes and, perhaps more importantly, these genes rarely undergo horizontal gene transfer and so are thought to be reliable for identifying truly new bugs out there. The problem has been that the process of taking a slew of these sequences and cranking them through available software to see who was living in your favorite type of sludge was a tedious process involving many different programs. The slowest part of this process was often manual editing of the matrix to adjust alignments. Recently, Wu et al. developed a new pipeline that can completely automate this process, called "Small Subunit rRNA Taxonomy and Alignment Pipeline" or STAP, for short. Although getting this pipeline going requires that you have a basic bioinformatics toolkit installed and compiled (including ClustalW, PhyML and some BioPerl scripts), once this thing is up and running, it is fast (& parallelizable), reliable (more reliable than BLASTN as you approach finer taxonomic scales - see figure above), and open source. Kudos to fellow blogger Jonathan Eisen and his crew for making the lives of Venter-ites everywhere a little easier - and bacterial taxonomy and systematics a lot more solid at the same time.
Tuesday, July 8, 2008
New SSU Alignment and Phylogenetic Pipeline: STAP
Ok, here's where I, as the "microbial" person on this blog start speaking another language: most microbiologists, especially those who conduct large-scale environmental sampling for novel lineages of bacteria, archaea, and microbial eukaryotes, still use (gasp!) small subunit ribosomal RNA sequences for identifying organisms and analyzing communities. The reason is simple - primers to amplify these genes are almost completely universal so can be counted on to pick out even rare ("unculturable" almost goes without saying) microbes and, perhaps more importantly, these genes rarely undergo horizontal gene transfer and so are thought to be reliable for identifying truly new bugs out there. The problem has been that the process of taking a slew of these sequences and cranking them through available software to see who was living in your favorite type of sludge was a tedious process involving many different programs. The slowest part of this process was often manual editing of the matrix to adjust alignments. Recently, Wu et al. developed a new pipeline that can completely automate this process, called "Small Subunit rRNA Taxonomy and Alignment Pipeline" or STAP, for short. Although getting this pipeline going requires that you have a basic bioinformatics toolkit installed and compiled (including ClustalW, PhyML and some BioPerl scripts), once this thing is up and running, it is fast (& parallelizable), reliable (more reliable than BLASTN as you approach finer taxonomic scales - see figure above), and open source. Kudos to fellow blogger Jonathan Eisen and his crew for making the lives of Venter-ites everywhere a little easier - and bacterial taxonomy and systematics a lot more solid at the same time.
Subscribe to:
Post Comments (Atom)
5 comments:
Note that this article is published in the PLoS title - PLoS One - that recently attracted the ire of Nature. True to form of course, you can't actually read this criticism unless you have payed access to the journal. Eisen, who is the academic editor for PLoS, has already responded over at his blog (including a catalogue of links to other responses in the blogosphere). There's no question that PLoS One is publishing some substandard work. I, for one, don't really care what PLoS has to do to succeed...as long as they keep Nature on the run. Do you really think all the big time publishers would be moving toward more open access as quickly as they are if there weren't a trouble maker like PLoS on the scene?
I really liked Eisen's discussion of the PLoS v. Nature issue.
It's obviously very important that the quality of publications - open access or not - stay high, but in a time when science literacy, respect, and funding is at a horrifying low, I'm definitely in favor of anything that can give people access to real research and scientific discussion, both. It shouldn't cost $50 to read a scientific paper. That's just plan wrong.
Hey Susan, here's a link for some additional discussion about STAP:
http://scienceblogs.com/mikethemadbiologist/2008/07/some_thoughts_about_stap.php
It may or may not be useful!
Thanks, Jess - thought it was particularly cool that Eisen addressed most of those points almost immediately!
Note - as I posted on Mike the Mad Biologist's blog that Jess links to --- we are trying to shift away from our proile alignment method that is used in STAP to using a secondary structure based alignment program that is under development by Sean Eddy.
And glor I definitely agree that, whatever you think of PLoS One or PLoS, there is no doubt that PLoS overall has helped to push the commercial and academic publishers towards OA.
Post a Comment