multiple_rarefaction_even_depth results in the same OTU table 100 times

[Alyse]

Hello, 

I am using macqiime 1.9.0 to perform rarefaction on a microbial dataset generated by Illumina MiSeq. I used multiple_rarefactions_even_depth.py to rarefy to a depth of 30,000 sequences with a total of 100 replicates. The command I used was: 

multiple_rarefactions_even_depth.py -i otutable.biom -o otu_tables_30000_100x/ -d 30000 -n 100 -k 

However, all of the 100 OTU tables are exactly the same (i.e., standard deviation = 0). This does not seem correct to me. Shouldn't there be some degree of variability between each OTU table?  

Thanks so much! 

Alyse 

[Jamie Morton]

Hi Alyse,

That is strange.  Would you like to post your biom table so I can reproduce the result?

Best,

Jamie

[Alyse]

Hi Jamie, 

I have attached my OTU table. Let me know if you get the same thing. 

Thank so much! 

Alyse 

[Jamie Morton]

Hi Alyse,

Running the following, I didn't see any such differences

multiple_rarefactions_even_depth.py -i otutable.biom -o 100x -d 100 -n 10
cd 100x

(qiime)tests-MacBook-Pro-3:100x mortonjt$ biom summarize-table -i rarefaction_100_0.biom --observations | tail
OTU_1681;size=915;: 75.0
OTU_278;size=1920;: 75.0
OTU_5359;size=16;: 84.0
OTU_247;size=6446;: 107.0
OTU_18;size=40505;: 112.0
OTU_28;size=34144;: 126.0
OTU_4999;size=14;: 385.0
OTU_5576;size=10;: 537.0
OTU_2;size=2268779;: 2517.0
OTU_1;size=3592356;: 6068.0
(qiime)tests-MacBook-Pro-3:100x mortonjt$ biom summarize-table -i rarefaction_100_1.biom --observations | tail
OTU_278;size=1920;: 64.0
OTU_1681;size=915;: 68.0
OTU_5359;size=16;: 88.0
OTU_18;size=40505;: 96.0
OTU_247;size=6446;: 103.0
OTU_28;size=34144;: 143.0
OTU_4999;size=14;: 368.0
OTU_5576;size=10;: 536.0
OTU_2;size=2268779;: 2611.0
OTU_1;size=3592356;: 6052.0

Do you see these sorts of results when you run biom summarize-table -i rarefaction_100_1.biom --observations ?

Jamie

[Alyse]

Hi Jamie, 

I ran biom summarize-table -i rarefaction_30000_x.biom --observation for the first five rarefactions and did indeed see variability in the total sequence count for each OTU. I attached a .csv here as evidence (please ignore the "size" in the OTU name - it's an artifact from UPARSE). 

Perhaps the reason why I saw standard deviations of 0 is because I took the standard deviation of each OTU at each station (i.e., sampling site) across all 100 replicate rarefactions? I'm thinking that the variability of each OTU at each station is very low, even though you do see variability in the summed sequence abundance for each OTU across all stations. 

Thanks again, 

Alyse