Regarding otu filter

[divyapr...@gmail.com]

Hi every body

since i have posted the same question in the previous thread but did not got the reply. so i a creating a new thread for the same question

 

. Being new to Qiime i am not able to do otu  filtering. i mean i don't know whether i have to run filter_otus_from_otu_table.py or   filter_samples_from_otu_table.py or filter_taxa_from_otu_table.py.

I have got three samples, with otu count 3412. what is the better option for me . So that i can proceed forward to prepare phylogenetic tree and heatmap and can do other analysis. 

More over i want to known after filtering with which file i have to proceed with.

 

[Jenya Kopylov]

Hello,

Have you read the documentation for all three scripts? Each documentation page includes examples you could follow.

http://qiime.org/scripts/filter_otus_from_otu_table.html

http://qiime.org/scripts/filter_samples_from_otu_table.html

http://qiime.org/scripts/filter_taxa_from_otu_table.html

Your question is difficult to answer unless we understand the reason why you want to filter your OTU table in the first place. What exactly do you want to remove from it?

For example, often users will want to remove singleton OTUs (OTUs with count = 1) from further analysis since these can be due to sequencing errors / noise. For this reason, you would want to use filter_otus_from_otu_table.py .

Jenya

[divyapr...@gmail.com]

Thanks Jenya 

I have read these. Actually what i want is that i am getting

1. i am getting otu_heatmap pdf. that is almost not visible. How to make it visible.

2. otu_network that too is more congested not able to site it properly. what i want is to view it properly and clearly 

. May be it may be done by filtering _otus.

3. Phylogenetic trees also generating large number of otus. not able to view it properly. and more over not able to convert_tree_tips-to taxas.

However i think it is not wise to remove otus. than what is the fun of doing sequencing.

i am attaching all the three files for your convenience 

[Jenya Kopylov]

Can you please paste exact commands used to generate each of the three PDFs you attached?

Thanks,

[divyapr...@gmail.com]

Hi Jenya

 command for otu_heatmap: make_otu_heatmap.py -i Soil_otus/otu_table.biom -o otu_heatmap

 command for otu_network: make_otu_network.py -i Soil_otus/otu_table.biom -m Soil_mapped_file/Soil_mapping.csv_corrected.txt -o otu_network and after the command is run i used cytoscape to develop network.

command for phylogenetic tree: after obtaining the pynast_aligned_seqs. and rep set  by running pick_de_novo_otus.py  and then typing figtree in the terminal and then developing Fig tree

[Kyle Bittinger]

Again, I think I addressed this question in another thread, please see the answer there.

In response to your comment about "removing OTUs," I think you are right that one generally wants to be careful about removing OTUs for analysis.  However, I think it is fine to select a smaller number of OTUs for visualization in a plot.

[divyapr...@gmail.com]

Thanks Jenya

please mention how can i select a smaller number of OTUs for visualization in a plot.

I got the point how to visualize the otu heatmap by simply increasing its height.

How ever i still have doubts regarding the otu_network, how to visualize that.

[roy1234...@gmail.com]

Hi divya 

i have also got the same issues with otu_network want to visualize it properly.

If any suggestions please paste it here

[Colin Brislawn]

Hello,

Like Kyle posted in the other thread, you can use this script to remove OTUs from a table:

http://qiime.org/scripts/filter_otus_from_otu_table.html 

That script has many different ways to filter, so when you write the figure caption, it's important to mention the settings you use. For example, if you use
filter_otus_from_otu_table.py -i otu_table.biom -o otu_table_more_than_3_samples.biom -n 3
for making a heatmap, you could write this in the figure caption:

"Heatmap showing OTUs that appeared in three or more samples..." 

Colin