How to calculate alpha diversity using many different metrics in one command?

User Support
1

Hello,

Does anyone know how to calculate alpha diversity using several different metrics in one command please?

I tried following the advice given on How to compute several alpha-diversity rarefaction plots - #3 by Nicholas_Bokulich, but it gave me an error.

My command:

qiime diversity alpha
--i-table Ar2_raw_PairedEnd_table_deblur.qza
--p-metric 'observed_otus'
--p-metric 'shannon'
--p-metric 'dominance'
--p-metric 'osd'
--p-metric 'chao1'
--p-metric 'strong'
--p-metric 'goods_coverage'
--p-metric 'lladser_pe'
--p-metric 'mcintosh_d'
--p-metric 'brillouin_d'
--p-metric 'ace'
--p-metric 'michaelis_menten_fit'
--p-metric 'enspie'
--p-metric 'fisher_alpha'
--p-metric 'kempton_taylor_q'
--p-metric 'lladser_ci'
--p-metric 'berger_parker_d'
--p-metric 'esty_ci'
--p-metric 'gini_index'
--p-metric 'simpson_e'
--p-metric 'doubles'
--p-metric 'menhinick'
--p-metric 'singles'
--p-metric 'heip_e'
--p-metric 'mcintosh_e'
--p-metric 'chao1_ci'
--p-metric 'margalef'
--p-metric 'pielou_e'
--p-metric 'simpson'
--p-metric 'robbins'
--o-alpha-diversity Ar2_raw_PairedEnd_alpha_diversity_observed_otus.qza

And this is the error I got:

              There were some problems with the command:

(1/10?) Option '--p-metric' was specified multiple times in the command.
(2/10?) Option '--p-metric' was specified multiple times in the command.
(3/10?) Option '--p-metric' was specified multiple times in the command.
(4/10?) Option '--p-metric' was specified multiple times in the command.
(5/10?) Option '--p-metric' was specified multiple times in the command.
(6/10?) Option '--p-metric' was specified multiple times in the command.
(7/10?) Option '--p-metric' was specified multiple times in the command.
(8/10?) Option '--p-metric' was specified multiple times in the command.
(9/10?) Option '--p-metric' was specified multiple times in the command.
(10/10?) Option '--p-metric' was specified multiple times in the command.

I would really appreciate your advice.. Many thanks!

2

Hi @fgara,

Welcome to the :qiime2: forum.

At the moment, it's not possible to do this in one command in QIIME 2. You could build a for-loop (I'd recommend the python API for that but if you have some bash-foo, you could probably put together a loop on the command line or via a snakemake pipeline. You could alos check and see if core-diversity will let you change the list of metrics.

Otherwise, this might be a good development suggestion.

Best,
Justine

3

Thank you!
I will try a for-loop in that case.
And yes, if it's possible to enable this as one command in QIIME2, that would be great!

Thanks again

4

Hi,

Sorry for asking again.

Following your suggestion I tried the following, but it gave me error.

for i in shannon observed_otus dominance
do
qiime diversity alpha
--i-table Ar2_raw_PairedEnd_table_deblur.qza
-–p-metric "'$i'"
--o-alpha-diversity Ar2_raw_PairedEnd_alpha_diversity
done

Error message:
Error: Detected invalid character in: -–p-metric
Verify the correct quotes or dashes (ASCII) are being used.

I have tried various combinations of quotes, such as "", '', "''",... all gave me the same error message.

Do you know how to correctly pass on a bash variable to a QIIME command please?

Thank you so much once again

5

HI @fgara,

I think you can just try

for i in shannon observed_otus dominance
do
qiime diversity alpha
–i-table Ar2_raw_PairedEnd_table_deblur.qza
-–p-metric $i
–o-alpha-diversity Ar2_raw_PairedEnd_alpha_diversity
done

But, bash is not my best language.

Best,
Justine

6

Dear Justine,

I've tried your suggestion, and still got the same error:
Error: Detected invalid character in: -–p-metric, –-o-alpha-diversity
Verify the correct quotes or dashes (ASCII) are being used.

But don't worry, I'll try to figure it out.
Thanks again for your kindness!

7

Hi @fgara,

Again, bash isn't my language, but maybe try

for i in 'shannon' 'observed_otus' 'dominance'
do
echo $i
qiime diversity alpha
--i-table Ar2_raw_PairedEnd_table_deblur.qza
--p-metric $i
--o-alpha-diversity $i.qza
done

I also think you want to direct your output path more clearly. Like, you'll re-write the filepath if you do it this way.

11

So, a wonderful collegue pointed out that you might also have an issue with your fancy dashes. Try typing directly on the command line or using a text editor to format

Best,
Justine

13

Hi, @fgara
I have some ideas for your problem.

If you have a metada file, you can try ‘qiime diversity core-metrics’ or ’ qiime diversity core-metrics-phylogenetic ’ to calculate alpha diversity, and you will get some of the alpha diversity you listed (see the command documentation for details).

There’s also a very silly way. Your code will print out the qiime2 commands. If you want q2 to run these commands, you can try the following:

Create a blank ’ .txt ’ file, write the qiime2 command with a different ’ --p-metric ’ parameter by line through the loop or manually, then, change the file suffix to ’ .sh ', like this:

qiime diversity --i-table table.qza --p-metric shannon --o-alpha-diversity output.qza
qiime diversity --i-table table.qza --p-metric observed_otus --o-alpha-diversity output.qza

and run it on the terminal using the ‘nohup sh your_file_name.sh &’ command.

Hope the above can help you.

14

Dear both,

Thanks so much for your kind replies.
Unfortunately, I've tried all sorts of different dashes, quotation marks etc but nothing worked.
So I decided to write all of the commands in a bash file script.

Many thanks once again