The bioBakery help forum

MetaPhlAn3 gives only bacterial profile

Hi
I tried using MetaPhlAn-3 for profiling my shotgun metagenome sequence from Nova-seq, I have few doubts and queries for utilizing the best of MetaPhlAn-3.

  1. MetaPhlAn can be used for environmental samples too or is it just for human samples only?
  2. I tried the basic command :
    metaphlan --bowtie2db ~/Documents/MetaPhlAn-3/MetaPhlAn3-DB/ ~/Documents/FLOOR-METAGENOME-NOVASEQ/CAT_FASTQ_FILES/M2_S60_R1_fastq.gz ~/Documents/FLOOR-METAGENOME-NOVASEQ/CAT_FASTQ_FILES/M2_S60_R2_fastq.gz --bowtie2out M2-S60.bowtie2.bz2 --nproc 12 --input_type fastq
    It gave me profile.txt file with just the bacterial information. But when I tried to blast my sequences and get taxonomy though MEGAN6 I am getting Bacterial, Archeal, Eukaryotes as well as few viral clades too.
    I used the latest database which is mpa_v30_CHOCOPhlAn_201901
  3. Is there any place where you have explained about the parameters which are used in metaphlan3?
  4. What can I do to get results for All domains(Eukaryotic, Archeal, Bacterial and viral) as you have mentioned in your introduction about metaphlan3?

Hi,

  1. MetaPhlAn is also suitable for non-human samples, althrough, the profiling can be partial due to lack of the markers for the species, we expanded the database and included more species, even environmental-associated.
  2. The command you ran is not correct. If you have multiple files to profile, the input list should be comma separated, without spaces. I can suggest you to run the analysis running MetaPhlAn using --stat_q 0.1 to see if you have more support for more species.
  3. Yes, the parameters are explained in the tutorial, wiki, and in the help message (metaphlan -h)
  4. If enough reads are mapping to the markers for a species, no matter its domain, the relative abundance will be reported in the profile. Species not listed in the profile can be missing due to lack of support of the markers (too few) or the species is absent in the database.