Hope you’re doing well! PICRUSt is a pretty neat tool! It seems very meaningful in terms of complementing our current data analysis plan. One highlight is the way in which it would allow us to elucidate the roles each OTU plays in differential gene functioning. For the metagenome inference step, it seems that we can produce a table of OTUs, specific to each sample, using QIIME. This can then be computationally normalized, giving us as a final output the predicted gene family counts for each sample. In my opinion, this would be pretty significant because by allowing us to estimate the relative abundance of 16S rRNA gene families within the community, we can perhaps infer the ways in which these differential gene families interact to impact the healing process in SCD leg ulcers. It would also shed qualitative light on the differences between the microbial communities of the cohort groups. A few questions I had: Would we use the installed PICRUSt or the Galaxy platform? How do we determine the identity percentage we’d like to pick our OTUs at? There seems to be quite a few directions we can go in with the software. I can also see why coding is important, haha. I’ve been dabbling in Python, via Codecademy, but it has only taught me the basics. I’m looking for another resource at the moment!
With respect to the Shannon Diversity index, I’ve learned that, in contrast to other indices, it takes into account both richness, the number of species present, and evenness, their relative abundances. I found these two articles particularly insightful: http://f1000.com/prime/reports/b/6/51/http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0032118
The first one underscores some of the weaknesses that accompany PICRUSt, which I thought might be useful to know (If you don't have much time, I would check out the PICRUSt subheading, but the whole article is worth a read, in my opinion). The latter journal article examines the way in which the Shannon index may be employed, and also provides some warnings to be heedful of. For example, the skin tends to contain large amounts of low abundant taxa that the Shannon index often fails to take into account, which may lead to a faulty inference regarding the relative diversity when comparing two different microbial communities. I discovered that it may be critical to employ the T statistic (or to make a species area curve assessment) in order to circumvent some of the index’s inherent weaknesses.
@katebliz
Hi Kate,
Hope you’re doing well! PICRUSt is a pretty neat tool! It seems very meaningful in terms of complementing our current data analysis plan. One highlight is the way in which it would allow us to elucidate the roles each OTU plays in differential gene functioning. For the metagenome inference step, it seems that we can produce a table of OTUs, specific to each sample, using QIIME. This can then be computationally normalized, giving us as a final output the predicted gene family counts for each sample. In my opinion, this would be pretty significant because by allowing us to estimate the relative abundance of 16S rRNA gene families within the community, we can perhaps infer the ways in which these differential gene families interact to impact the healing process in SCD leg ulcers. It would also shed qualitative light on the differences between the microbial communities of the cohort groups. A few questions I had: Would we use the installed PICRUSt or the Galaxy platform? How do we determine the identity percentage we’d like to pick our OTUs at? There seems to be quite a few directions we can go in with the software. I can also see why coding is important, haha. I’ve been dabbling in Python, via Codecademy, but it has only taught me the basics. I’m looking for another resource at the moment!
With respect to the Shannon Diversity index, I’ve learned that, in contrast to other indices, it takes into account both richness, the number of species present, and evenness, their relative abundances. I found these two articles particularly insightful: http://f1000.com/prime/reports/b/6/51/ http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0032118 The first one underscores some of the weaknesses that accompany PICRUSt, which I thought might be useful to know (If you don't have much time, I would check out the PICRUSt subheading, but the whole article is worth a read, in my opinion). The latter journal article examines the way in which the Shannon index may be employed, and also provides some warnings to be heedful of. For example, the skin tends to contain large amounts of low abundant taxa that the Shannon index often fails to take into account, which may lead to a faulty inference regarding the relative diversity when comparing two different microbial communities. I discovered that it may be critical to employ the T statistic (or to make a species area curve assessment) in order to circumvent some of the index’s inherent weaknesses.
-Anitra