Can calour detect multipile groups differential abundance?

[biofuture]

Hi,

I would like to know whether this tool can be used to identify multiple groups differential abundance microbes?

Thank you very much!

[amnona]

Currently there is no statistical test for multiple group difference in calour :(

1. I usually like to look at the data (all categories) and then do tests on groups that look different (sometimes joining several groups into one group for the test). Usually i believe that if you don't see a difference first by eye in the heatmap, it is not that biologically interesting...
2. Will be happy if you want to contribute code for multiple group differential abundance testing :). Note that currently the mutiple hypothesis correction method is dsFDR which is based on permutations, so ideally to incorporate the test into that framework (but can also have it just use BH-FDR instead of too slow)

[biofuture]

thanks，then I can just use the qiime2 ds_fdr plugin for multiple testing? Is that right?

[amnona]

In theory yes. But since dsFDR is based on permutations, you need the statistic calculation to be fast enough (since it needs to calculate it for 1000 random permutations). What test exactly do you want to do? what commands did you plan to use from ds_fdr plugin?

On Tue, Oct 16, 2018 at 2:06 PM Xiaotao JIANG (姜小濤) < notifications@github.com> wrote:

thanks，then I can just use the qiime2 ds_fdr plugin for multiple testing? Is that right?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/biocore/calour/issues/136#issuecomment-430195910, or mute the thread https://github.com/notifications/unsubscribe-auth/AFkA8przkbNJA4cPb0JHj3ethdw8TM6oks5ulb2bgaJpZM4XdrXo .

[biofuture]

Is this is the correct way to use the tool?

qiime dsfdr permutation-fdr Usage: qiime dsfdr permutation-fdr [OPTIONS]

Discrete FDR

Options: --i-table ARTIFACT PATH FeatureTable[Frequency] [required] --m-metadata-file MULTIPLE PATH Metadata file or artifact viewable as metadata. This option may be supplied multiple times to merge metadata. [required] --m-metadata-column MetadataColumn[Categorical] Column from metadata file or artifact viewable as metadata. [required] --p-statistical-test [pearson|nonzerospearman|spearman|meandiff|stdmeandiff|nonzeropearson|mannwhiteny|kruwallis] [default: meandiff] --p-transform-function [rank|pa|norm|log] [default: rank] --p-alpha FLOAT [default: 0.05] --p-permutations INTEGER [default: 1000] --o-visualization VISUALIZATION PATH [required if not passing --output-dir] --output-dir DIRECTORY Output unspecified results to a directory --cmd-config PATH Use config file for command options --verbose Display verbose output to stdout and/or stderr during execution of this action. [default: False] --quiet Silence output if execution is successful (silence is golden). [default: False] --citations Show citations and exit. --help Show this message and exit.

Error: Missing option: --i-table Error: Missing option: --m-metadata-file Error: Missing option: --m-metadata-column Error: Missing option: --o-visualization Note: When only providing names for a subset of the output Artifacts or Visualizations, you must specify an output directory through use of the --output-dir DIRECTORY flag.

y yes. But since dsFDR is based on permutations, you need the

statistic calculation to be fast enough (since it needs to calculate it for 1000 random permutations). What test exactly do you want to do? what commands did you plan to use from ds_fdr plugin?

On Tue, Oct 16, 2018 at 2:06 PM Xiaotao JIANG (姜小濤) < notifications@github.com> wrote:

thanks，then I can just use the qiime2 ds_fdr plugin for multiple testing? Is that right?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/biocore/calour/issues/136#issuecomment-430195910, or mute the thread < https://github.com/notifications/unsubscribe-auth/AFkA8przkbNJA4cPb0JHj3ethdw8TM6oks5ulb2bgaJpZM4XdrXo

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/biocore/calour/issues/136#issuecomment-430199555, or mute the thread https://github.com/notifications/unsubscribe-auth/ABaxoM9zBkIwa3bbhhehZyB-penKsoBjks5ulcDPgaJpZM4XdrXo .

[amnona]

i don't have experience with the qiime2 plugin, and especially with the kruwallis test. Let me know how it goes, and if you encounter problems, you can post a question in the dsfdr github repo. Again, i recommend to first look at the heatmap in calour, and see what groups show a difference in bacteria (by zooming in and observing by eye). How many groups do you have? what are they?

On Tue, Oct 16, 2018 at 3:19 PM Xiaotao JIANG (姜小濤) < notifications@github.com> wrote:

Is this is the correct way to use the tool?

qiime dsfdr permutation-fdr Usage: qiime dsfdr permutation-fdr [OPTIONS]

Discrete FDR

Options: --i-table ARTIFACT PATH FeatureTable[Frequency] [required] --m-metadata-file MULTIPLE PATH Metadata file or artifact viewable as metadata. This option may be supplied multiple times to merge metadata. [required] --m-metadata-column MetadataColumn[Categorical] Column from metadata file or artifact viewable as metadata. [required] --p-statistical-test

[pearson|nonzerospearman|spearman|meandiff|stdmeandiff|nonzeropearson|mannwhiteny|kruwallis] [default: meandiff] --p-transform-function [rank|pa|norm|log] [default: rank] --p-alpha FLOAT [default: 0.05] --p-permutations INTEGER [default: 1000] --o-visualization VISUALIZATION PATH [required if not passing --output-dir] --output-dir DIRECTORY Output unspecified results to a directory --cmd-config PATH Use config file for command options --verbose Display verbose output to stdout and/or stderr during execution of this action. [default: False] --quiet Silence output if execution is successful (silence is golden). [default: False] --citations Show citations and exit. --help Show this message and exit.

Error: Missing option: --i-table Error: Missing option: --m-metadata-file Error: Missing option: --m-metadata-column Error: Missing option: --o-visualization Note: When only providing names for a subset of the output Artifacts or Visualizations, you must specify an output directory through use of the --output-dir DIRECTORY flag. Xiao-Tao Jiang, Ph.D. Postdoc Research Fellow Microbiome Research Centre St George and Sutherland Clinical School UNSW Sydney

Level 2, Clinical Sciences Building (WR Pitney) Short Street, St George Hospital KOGARAH NSW 2217 T: +61 402 943 681 Email: biofuture.jiang@gmail.com/xiaotao.jiang@unsw.edu.au MRC web: https://microbiome.org.au/

amnona notifications@github.com 于2018年10月16日周二 下午10:19写道：

In theory yes. But since dsFDR is based on permutations, you need the statistic calculation to be fast enough (since it needs to calculate it for 1000 random permutations). What test exactly do you want to do? what commands did you plan to use from ds_fdr plugin?

On Tue, Oct 16, 2018 at 2:06 PM Xiaotao JIANG (姜小濤) < notifications@github.com> wrote:

thanks，then I can just use the qiime2 ds_fdr plugin for multiple testing? Is that right?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/biocore/calour/issues/136#issuecomment-430195910, or mute the thread <

https://github.com/notifications/unsubscribe-auth/AFkA8przkbNJA4cPb0JHj3ethdw8TM6oks5ulb2bgaJpZM4XdrXo

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/biocore/calour/issues/136#issuecomment-430199555, or mute the thread < https://github.com/notifications/unsubscribe-auth/ABaxoM9zBkIwa3bbhhehZyB-penKsoBjks5ulcDPgaJpZM4XdrXo

.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/biocore/calour/issues/136#issuecomment-430215523, or mute the thread https://github.com/notifications/unsubscribe-auth/AFkA8nGhFdtmwpyEgctpR_4QuLvRQe4Cks5ulc7igaJpZM4XdrXo .