Moore Notes 6 24 09
1. Steve: would be useful to be able to root trees. Is there some outgroup that is included in the AMPHORA reference sequences? Is there some easy way we could get a single archaeal gene to root the tree? Might get long branch attraction problem. If really want rooted tree, should root at thermas or dynochoccus.
Steve trying to get RAXML working with monster alignment. He got it working. New version RAXML, algorithm for placing short sequences onto reference tree. Based on Maximum liklihood.
2. Sam: she is sampling reads from AMPHORA reference sequence data base. Breaking up into reads using metasim. using amphora to use trimmed alignment using AMPHORA hhm. Calling marker scanner script of amphora. It always dropping some of the reads.
3. Morgan: has been using Hummer version 3 (running GOS against pFAM) with genbeo, worked in 4 days.
4. Katie: talking about two things: 1) how to deal with running a huge amount of data, Martin has solution for that, 2) working with simulator, should have reads dropped. I don't know if need first BLAST step.
5. Katie's paraphrase: within a certain group of species, there may be insertion. If that is true, insertion won't show up in profile. Will be low quality read within the profile. Masking at alignment level. It is just a mark. Trimming also at alignment level, but actually removing.
6. Katie: Martin if it would be possible documenting this, it would be really useful. Possible to put a few comments in the script? He already put some info in the readme file, but Katie would like some more.
7. Josh posted a data request on openwetware. Srijak ordered the new data request set up (using project management software). Should arrive in a week. Had anyone created OTUs based on % similarity using 16S in PLoS data. It seems like must have been done already, but can't find publication talks about it.
8. Steve: do we have 16S alignment for all sites? Donying doesn't think there are already alignment files. Martin thinks can just do blast search against GOS.
9. Jess: check out Quince et al. 2008
10. Donying: Building gene families for 100 genomes across bacteria and archaea. Identify those families with single copy genes across all species. Use topology comparison to compare trees. Found 25 additional marker candidates for amphora. Still doing this for actino. Problem now is topology comparison not good, trees each nodes and edge equally. Get low comparative score even if looks good. Still don't know how to take different gene families, and normalize.
11. Josh: think it would be worthwhile to discuss what we mean about correspondence level of similarities.
12. Katie: will have weights that tell you where are in the tree, where on gene you are, etc. Won't be perfect, but come up with proposal.
13. Annual meeting: he is going to e-mail Kelly about alternative dates.
14. Josh posted Harte al. SAR paper.