Wow, Dan! Really nice video! One question, why is the standard deviation used in the rarefaction curve and not other error bars, for example, confidence intervals?
I have to make a bray curtis matrix by hand for a dataset of 3 sites with 4 species each. Site one for each of 4 species, the counts are as follows: 20 , 10, 6, 0 Site two: 5, 0, 2, 10. For site three: 0, 13, 3 20. by my understandig, since the minimum count for each species at each site is zero, so the sum of minima is zero. you can't divide anything by zero, so the matrix would be all zeros.
So Bray-Curtis gives you an average of all the low counts in two samples. That would mean that if at least one of the two samples has mostly low values, the BC diversity will be low. That would be the case even if both samples have the exact same values for all species, but all of them are low. Which does not make much sense, because the diversity there would be zero or null.
Not in QIIME. Anything which has been developed for macro-ecology can in principle also used for microbiome data, but with mountains of salt. So, for beta partition, R-package betapart is helpful. However, in a classical NGS dataset (count data, quite possible zero-inflated, compositional and so on) with blurry OTU definitions and even blurrier taxonomic resolution, and ~90% rare OTUs, i would really be careful with terms like nestedness and turnover.
First of all, thank you for sharing a priceless video. I have excel data come from Pacbio platform. Can I make a beta diversity without metadata? is there an application you recommend for beginners like me. thank you
That biom package is not available !! Coule you please check and let me know !! I have used below link for downloading . metagenome.cs.umn.edu/microbiomecodebrowser/doc/loading.data.into.R.html
So it appears that the 'biom' package has been deprecated. you can still use the biom format by installing bioconductor. Run the following script in R: ## try if URLs are not supported source("bioconductor.org/biocLite.R") biocLite("biomformat") ##end of command Source: bioconductor.org/packages/release/bioc/html/biomformat.html Happy sequencing!
Can't focus on the video, Dan. You're such a stud
My goodness this is such a relief to find. Thank you!
Great video, it has been really helpful.
Wow, Dan! Really nice video! One question, why is the standard deviation used in the rarefaction curve and not other error bars, for example, confidence intervals?
What are the % in the axis, what decides them and how can we interpret them? Thank you in advance
I have to make a bray curtis matrix by hand for a dataset of 3 sites with 4 species each. Site one for each of 4 species, the counts are as follows: 20 , 10, 6, 0
Site two: 5, 0, 2, 10. For site three: 0, 13, 3 20. by my understandig, since the minimum count for each species at each site is zero, so the sum of minima is zero. you can't divide anything by zero, so the matrix would be all zeros.
Thank you.
Awesome
The old horseshoe effect! PCA and PCoA probably aren't great options for ordinating this dataset.
So Bray-Curtis gives you an average of all the low counts in two samples. That would mean that if at least one of the two samples has mostly low values, the BC diversity will be low. That would be the case even if both samples have the exact same values for all species, but all of them are low. Which does not make much sense, because the diversity there would be zero or null.
Helpful.
I have a question, what about beta partition?
How much is nested? shared? diferenced in terms of richness?
Not in QIIME. Anything which has been developed for macro-ecology can in principle also used for microbiome data, but with mountains of salt. So, for beta partition, R-package betapart is helpful. However, in a classical NGS dataset (count data, quite possible zero-inflated, compositional and so on) with blurry OTU definitions and even blurrier taxonomic resolution, and ~90% rare OTUs, i would really be careful with terms like nestedness and turnover.
First of all, thank you for sharing a priceless video. I have excel data come from Pacbio platform. Can I make a beta diversity without metadata? is there an application you recommend for beginners like me.
thank you
how to calculate B diversity and importance to calculate it ?
share vedios ??
I think it must be updated because rarefaction is not a good practice
That biom package is not available !! Coule you please check and let me know !! I have used below link for downloading . metagenome.cs.umn.edu/microbiomecodebrowser/doc/loading.data.into.R.html
So it appears that the 'biom' package has been deprecated. you can still use the biom format by installing bioconductor. Run the following script in R:
## try if URLs are not supported
source("bioconductor.org/biocLite.R")
biocLite("biomformat")
##end of command
Source: bioconductor.org/packages/release/bioc/html/biomformat.html
Happy sequencing!
Thank you so much
This is hard to understand for a QIIME2 user