search

biocore / qiime

Official QIIME 1 software repository. QIIME 2 (https://qiime2.org) has succeeded QIIME 1 as of January 2018.
GNU General Public License v2.0
286 stars 265 forks source link

Document and add references for beta diversity metrics #1939

Open adamrp opened 9 years ago

adamrp commented 9 years ago

AFAIK, the documentation requested by the author of this post on the QIIME forum does not exist.

adamrp commented 9 years ago

This is a comment from @wdwvt1 in that QIIME forum thread. Copying it here since it provides the relevant reference/information for one of the metrics:

asymmetric UniFrac is also known as UniFrac G, gain, and Faith's Phylogenetic Diversity (its frustrating how many names it has). It is equivalent to Faith's Phylogenetic Diversity measure from his 1992 paper 'Conservation Evaluation and Phylogenetic Diversity' Daniel P. Faith, Biological Conservation 1992, 61, 1-10.

rob-knight commented 9 years ago

Not quite accurate. Faith's PD is alpha diversity in one sample, not a calculated difference between samples. The term asymmetric UniFrac should not be used: although we invented it independently, G (for Gain in PD) per Faith 1992 has priority and should be used.

On Feb 23, 2015, at 1:12 PM, adamrp notifications@github.com<mailto:notifications@github.com> wrote:

This is a comment from @wdwvt1https://github.com/wdwvt1 in that QIIME forum thread. Copying it here since it provides the relevant reference/information for one of the metrics:

asymmetric UniFrac is also known as UniFrac G, gain, and Faith's Phylogenetic Diversity (its frustrating how many names it has). It is equivalent to Faith's Phylogenetic Diversity measure from his 1992 paper 'Conservation Evaluation and Phylogenetic Diversity' Daniel P. Faith, Biological Conservation 1992, 61, 1-10.

— Reply to this email directly or view it on GitHubhttps://github.com/biocore/qiime/issues/1939#issuecomment-75633678.

adamrp commented 9 years ago

Thanks for the correction!

On Mon, Feb 23, 2015 at 2:38 PM, Rob Knight notifications@github.com wrote:

Not quite accurate. Faith's PD is alpha diversity in one sample, not a calculated difference between samples. The term asymmetric UniFrac should not be used: although we invented it independently, G (for Gain in PD) per Faith 1992 has priority and should be used.

On Feb 23, 2015, at 1:12 PM, adamrp <notifications@github.com<mailto: notifications@github.com>> wrote:

This is a comment from @wdwvt1https://github.com/wdwvt1 in that QIIME forum thread. Copying it here since it provides the relevant reference/information for one of the metrics:

asymmetric UniFrac is also known as UniFrac G, gain, and Faith's Phylogenetic Diversity (its frustrating how many names it has). It is equivalent to Faith's Phylogenetic Diversity measure from his 1992 paper 'Conservation Evaluation and Phylogenetic Diversity' Daniel P. Faith, Biological Conservation 1992, 61, 1-10.

— Reply to this email directly or view it on GitHub< https://github.com/biocore/qiime/issues/1939#issuecomment-75633678>.

— Reply to this email directly or view it on GitHub https://github.com/biocore/qiime/issues/1939#issuecomment-75639918.

ashleyshade commented 9 years ago

I think the original QIIME forum post was searching for an index of the specific mathematical definitions (or references to them) of all of the comparative diversity metrics. I am searching for the same information - how does eucbray_curtis differ mathematically from bray_curtis_magurran and bray_curtis_faith? Is there an issue open about this that I am missing? I want to use the "original" bray curtis metric.

antgonza commented 9 years ago

All the metrics can be found here and each of the methods have a reference and description. I think these are the lines you are looking for: l1, l2.

ashleyshade commented 9 years ago

Thank you!

Ashley Shade, Ph.D. Michigan State University Microbiology and Molecular Genetics Biomedical Physical Sciences rm 6179 517-884-5399 tel:517-884-5399 ShadeLab http://ashley17061.wix.com/shadelab

On Jun 18, 2015, at 4:17 PM, Antonio Gonzalez notifications@github.com wrote:

All the metrics can be found here https://github.com/pycogent/pycogent/blob/master/cogent/maths/distance_transform.py and each of the methods have a reference and description. I think these are the lines you are looking for: l1 https://github.com/pycogent/pycogent/blob/master/cogent/maths/distance_transform.py#L161, l2 https://github.com/pycogent/pycogent/blob/master/cogent/maths/distance_transform.py#L214.

— Reply to this email directly or view it on GitHub https://github.com/biocore/qiime/issues/1939#issuecomment-113277610.

ctekellogg commented 8 years ago

Hi there, I just wanted to follow up on this thread and see if any progress has been made towards documentation for the various beta diversity metrics (like http://scikit-bio.org/docs/latest/generated/skbio.diversity.alpha.html but for beta diversity). As far as I can tell only 2 metrics have been added to a similar page for beta diversity (http://scikit-bio.org/docs/latest/generated/skbio.diversity.beta.html). It would be very helpful to know the differences between the various unifrac methods implemented in QIIME (v. 1.9).

Thank you so much!!

colleen

colinbrislawn commented 7 years ago

Hello Colleen,

Because scikit-bio only adds the two unifrac distances, it only documents those. Other measures and matrices are imported from scipy.spatial.distance.pdist and are documented over there.

I could not find any direct documentation for UniFrac gain / Gain in PD. Is this implemented in Qiime 1.9.1?