Friday, October 23, 2009

Why I think a Phylogenetic Standard might not work

Currently we have many different formats for phylogenies (e.g., nexus, newick, extend newick, phylip, xml) and there is no doubt that putting in place a Phylogenetic Standard would be a good thing but just because a group of researchers get together and decide that a Phylogenetic Standard would be good doesn't make it happen. Unless the phylogenetic community is behind the decisions being made which is unlikely to happen any time soon as phylogeneticist already have a poor record of submitting their phylogenies to databases like TreeBase.
These are obviously points that you are aware of as you mention them in your statement. I only remark that unless publishers enforce a particular format and submission of the phylogenetic data into repositories (like GenBank/EMBL), then however noble the idea of a Phylogenetic Standard is, it is unlikely to be put into practice by phylogeneticists.


hlapp said...

Joseph: The TDWG Interest Group on Phylogenetic Standards (note the plural!) is an interest group, not a standard, let alone The One Standard. (Have you actually read the charter?) It's a group of researchers that got together and decided that commonly accepted and sufficiently expressive standards are both currently lacking and desirable for phylogenetic data. If it isn't practicing phylogeneticists and others working with phylogenetic data that make the necessary standards happen, including their adoption and ultimately enforcement, who else would be more likely in your opinion? Moreover, taxonomic standards have been put into practice in precisely this way by taxonomists. What makes you conclude that phylogeneticists cannot succeed in a likewise effort? Especially as the TDWG Interest Group is a community-driven initiative that is open for anyone (including dissenters!) to join and contribute to constructively. As for the poor record of standards compliance among phylogeneticists, isn't that much more likely a reflection of a lack of agreed upon standards that meet the needs of researchers, rather than an unwillingness to follow them? Aside from that, submission to data repositories is an issue that's rather separate from the compliance with or acceptance of data exchange standards within a community - just ask any of the GenBank curators (or people like Rod Page who are trying to reuse the data) for how consistent sequence metadata and annotations are across sequence submissions. Finally, if the absence of a common standard were indeed one of the main reasons that publishers can't enforce data submission to repositories, how does that argue for the initiative of putting one in place to be destined for failure?

Nico & Hilmar

blJOg said...

A chicken and egg story
It seems to me that the cart is being put before the horse. I think that a much more productive use of your time would be to get publishers to enforce submission of phylogenies to a database like TreeBase in the same way as they request the submission of sequence data to GenBank. TreeBase already has a standard format for submission (albeit a bit dated) so you wouldn't need to invent a new one. Although GenBank existed before the publishers enforced submission to it for publication, I believe it is the pressure from editors to submit that makes GenBank what it is today. Sadly there has been no such enforcement for phylogenies and this is NOT because of a lack useful standards.

ncellinese said...

Joseph: You are skirting the questions raised above. Your opinion that publishers should require and enforce data submission to repositories is appreciated - it is in fact one of the major objectives of Dryad and its partner journals, so rest assured that it will happen. The standard format for TreeBASE submission (you are presumably referring to NEXUS?) is not only a bit dated but simply not expressive enough to represent all the metadata in a predictable manner that people want to attach meanwhile. Ever heard about the multitudes of incompatible NEXUS "flavors"? Therefore, right now the cart you put in front of your horse is not a very good one. We one to build a good cart, in fact many, and the best place to do this is within a community setting driven by people like you and us. The mailing list of the Interest Group is open for anyone to join, and you'd be more than welcome to engage with us constructively.

Nico & Hilmar

Disqus for Evo-Karma