The purpose of this program is to quickly collect data from Phylosift (GitHub link) runs on multiple bins and collate into a single table.
The expectation of the input is that you have a folder containing all the Phylosift output folders for each bin. The program looks for the Krona html file and pulls the results from this file. I haven't found the other outfiles from Phylosift to be quite as informative as the Krona file, which is why I chose it.
The only dependencies are a couple of Perl modules I regularly use. I think these are actually available in the Perl download, so not sure you really need to install anything. The modules are: strict & Getopt::Std.
Download the repository from GitHub: Phylosift_to_table
After installation, pull up the help information:
perl phylosift_krona_to_table.pl -h
This will give you:
Help called:
Options:
-p = path to folder containing folders of phylosift results
-x = folder suffix
-r = folder prefix
-h = This help message
I have an input folder that looks like this:
phylosift_results (folder/directory)
├── S1_bin1_out (folder/directory)
│ ├── S1_bin1.concat.jnlp
│ ├── S1_bin1.html
│ ├── S1_bin1.jplace
│ ├── S1_bin1.xml
│ ├── alignDir
│ ├── blastDir
│ ├── marker_summary.txt
│ ├── run_info.txt
│ ├── sequence_taxa.1.txt
│ ├── sequence_taxa.txt
│ ├── sequence_taxa_summary.1.txt
│ ├── sequence_taxa_summary.txt
│ ├── taxa_90pct_HPD.txt
│ ├── taxasummary.txt
│ └── treeDir
├── S1_bin2_out (folder/directory)
│ ├── S1_bin2.concat.jnlp
│ ├── S1_bin2.html
│ ├── S1_bin2.jplace
│ ├── S1_bin2.xml
│ ├── alignDir
│ ├── blastDir
│ ├── marker_summary.txt
│ ├── run_info.txt
│ ├── sequence_taxa.1.txt
│ ├── sequence_taxa.txt
│ ├── sequence_taxa_summary.1.txt
│ ├── sequence_taxa_summary.txt
│ ├── taxa_90pct_HPD.txt
│ ├── taxasummary.txt
│ └── treeDir
├── etc...
Thus phylosift_results is a folder containing all the phylosift output folders from all my bins; that folder is what you should put for -p.
S1_bin1_out and S1_bin2_out each have _out as a suffix to the bin name. That suffix is what is provided to -x. If you have decided to name your Phylosift output folders with a prefix, such as phylosift_S1_bin1, you can provide phylosift_ to the -r option. You cannot have both a suffix and prefix.
Example command:
phylosift_krona_to_table.pl -p ./phylosift_results -x _out > S1_phylosift_resultstable.txt
This generates a table following some rules. Primarily, I'm interested in using this script to call a taxonomy for a bin, so I am looking for a majority rule. But I don't want to be too sloppy, so there are rules for what means a majority at different taxonomic levels. Multiple majority rule taxa are possible (and separated by ",") if they pass the majority cutoff for the taxonomic level.
| Taxonomic Level | Majority Cutoff | If below cutoff... |
|---|---|---|
| NULL | NA | WARNING: NO ASSIGNMENTS for bin |
| ROOT | 100% | WARNING: ROOT is not assigned to 100% for bin |
| Cellular organism | 100% | WARNING: Cellular organisms is not assigned to 100% for bin |
| Domain | 90% | Unclassified microbe |
| Phylum | 50% | Domain_name (mixed) |
| Class | 40% | Phylum_name (abund) [multi] |
| Order | 30% | Phylum_name (abund), Class_name (abund) [multi] |
| Family | 30% | Phylum/Class/Order (abund) [multi] |
| Genus | 25% | Phylum/Class/Order/Family (abund) [multi] |
| Species | 25% | P/C/O/F/G (abund) [multi] |
| Subspecies/Genome | 25% | P/C/O/F/G/S (abund) [multi] |
| Additional depths | 25% | P/C/O/F/G/S/Sub or Genome (abund) [multi] |
The output for sample Sample 1 looks like:
| Bin name | Taxonomy (percent relative abundance) | ||||||
|---|---|---|---|---|---|---|---|
| S1_bin1 | PROTEOBACTERIA (100) | GAMMAPROTEOBACTERIA (43.9969306483449) | |||||
| S1_bin2 | BACTEROIDETES/CHLOROBI GROUP (100) | BACTEROIDETES (100) | FLAVOBACTERIIA (100) | FLAVOBACTERIALES (57.1428571428571) | FLAVOBACTERIACEAE (44.092230634174) | ||
| S1_bin3 | PROTEOBACTERIA (100) | ZETAPROTEOBACTERIA (99.89503224484) | UNCLASSIFIED ZETAPROTEOBACTERIA (86.8889596220667) | ZETA PROTEOBACTERIUM SCGC AB 133 C04 (48.0392969962402),ZETA PROTEOBACTERIUM SCGC AB 137 I08 (36.5954238775949), | |||
| S1_bin4 | ARCHAEA (mixed) | ||||||
| WARNING: NO ASSIGNMENTS for bin S1_bin5 | |||||||
| S1_bin5 | BACTERIA (mixed) | ||||||
| S1_bin6 | Unclassified microbe | ||||||
| WARNING: Cellular organisms is not assigned to 100% for bin S1_bin7 | |||||||
| S1_bin8 | PLANCTOMYCETES (50) | CANDIDATE DIVISION NC10 (48.0386905602096),PHYCISPHAERAE (50), | |||||
| S1_bin9 | PROTEOBACTERIA (100) | GAMMAPROTEOBACTERIA (50),ZETAPROTEOBACTERIA (50), | |||||
| S1_bin10 | PROTEOBACTERIA (75.4290657552175) | DELTA/EPSILON SUBDIVISIONS (66.3622131607418) | DELTAPROTEOBACTERIA (66.3622131607418) | MYXOCOCCALES (49.1954721381515) | SORANGIINEAE (28.8032419805045) | POLYANGIACEAE (28.8032419805045) | SORANGIUM (28.8032419805045) |
In this example, bins 2/3/10 are a bit more specific, while bins 1/4/5/6 are not highly resolved by Phylosift. Bin 9 is split 50:50 between Zetaproteobacteria and Gammaproteobacteria. And bins 5/7 seem to have problems. Potentially phylosift didn't run properly on bin 5 (or didn't have any marker genes). And for bin 7, I tend to get this result when there are some viral hits in the Phylosift results (so should take a look at this manually to judge).
Please open an issue if you have any issues or questions. This program is provided without warranty or a guarantee of support. Thanks!!