question about the taxonomy result

[quliping]

Hellow, I used the "checkm lineage_wf" command and I'm confused about the "Marker lineage" column in the result. For example, the result of one of my bin is "s__algicola (UID2846)", but there are many species name that we call them "algicola" just like 'Bacillus algicola' and 'Crocinitomix algicola', so how should I know which kind of algicola my bin is?

[donovan-h-parks]

Hello. I do not recommend using CheckM for taxonomic classification. This is not the intended use of the software. You can determine the full taxonomic assignment of CheckM using the tree_qa command, e.g.:

checkm tree_qa -o2

https://github.com/Ecogenomics/CheckM/wiki/Genome-Quality-Commands#tree_qa

Cheers, Donovan

[quliping]

Thank you a lot! I konw what you mean and the checkM taxonomic result is just one of my reference. But there is still a problem that the tree_qa results are different from the Marker lineage column. Just like bin108, the tree_qa result is 'kBacteria;pActinobacteria;cActinobacteria', but the marker lineage result is 'salgicola'...The other one example, bin109, the results are 'kArchaea (root)' and 'kBacteria'... Also bin 113, they are 'kBacteria;pProteobacteria;cAlphaproteobacteria;oRhodospirillales;fRhodospirillaceae' and 'cBetaproteobacteria'... Of course I know they are different principles, one is base on genome tree the other is marker set. So the question is which kind of result should I trust? Or I should use another software just like PhyloPhlAn? I'm confused right now.

Thank you a lot!

Qu Liping

[donovan-h-parks]

Hello Qu. These results don't make sense to me. Both the phylogenetic placement and marker set used for completeness/contamination estimates should be in line with each other. Can you send me the relevant CheckM output files?

[quliping]

Of course I can. The file 'checkM_result_7_8.txt' is the result of 'lineage_wf' command and the 'tree_qa_result_7_8.txt' is the result of 'tree_qa' command. I also upload the genome tree from tree_qa which named 'checkm7_8.zip'. checkM_result_7_8.txt tree_qa_result_7_8.txt checkm7_8.zip

Thank you again.

Qu Liping

[donovan-h-parks]

Hello. Those results do not make sense to me. Both files use the same tree placement so it shouldn't be possible to get incongruent results. Can you send me the 3 bins you have flagged (108, 109, 113)?

[quliping]

OK. I will send the three bins to you. And I also put the other bins that have the same situation into the zip file. different_bins.zip And here are my commands: checkm lineage_wf -x fa -t 8 --pplacer_threads 8 bin_folder output_folder checkm tree_qa -o 2 -f checkm7_8.txt checkM7_8

I'm sorry to bother you so much. Thank you again.

[donovan-h-parks]

Hello. I'm not able to reproduce your results. The 108 and 113 bins use the kBacteria (UID203) marker set, while 109 uses the kArchaea (UID2) marker set. The tree_qa -o2 table indicates that: bin 108 = kBacteria;pActinobacteria;cActinobacteria bin 113 = kBacteria;pProteobacteria;cAlphaproteobacteria;oRhodospirillales;fRhodospirillaceae bin 109 = k__Archaea

This is all consistent and as I would expect. A marker set higher in the tree to where a bin is placed is often used as it was determined that this marker set likely gives a better estimate. For example, there many not be enough f__Rhodospirillaceae reference genomes to establish a reliable set of markers for estimating genome completeness and contamination.

[quliping]

I'm so sory... I went back to confirm the results. And I find something wrong with me. Actually there are two samples and both of them have 118 bins, so I think I I confused their results... I find the really results from the two commands that belong to sample7_8 are the same... So I made a mistake, not the software. I'm so sory, maybe you should close this question...

[donovan-h-parks]

Glad to hear the issue is resolved. Best of luck with your research.