Quality Check before metaphlan analysis

Microbial community profiling MetaPhlAn
1

Hi friends!!!
I want to analyse already submitted data in SRA/EMBL by MetaPhlAn3. Should I do quality check each time before I run with MetaPhlAn?

Thanks & Regards,
DC7

2

Hi, assuming that the submitted data was already QC’d, there’s no need to to QC.
Internally, MetaPhlAn discards reads longer less than 70bp

1 Like
3

Thanks @fbeghini. Does MetPhlAn discards sequences containing bases with low quality score?

DC7

4

Yes, marker hits with MAPQ below 5 are discarded

1 Like
5

But, Sir @fbeghini ,
how can I be sure that those submitted data were QC’d ?

6

If nothing is specified on SRA, the trimming + filtering of low MAPQ alignments done by MetaPhlAn is sufficient

1 Like
7

I have a WGS file where average read length is 35bp. So, are you saying that MetPhlAn will be unable to profile it. Right? (PRJEB23292)

8

No, you can set the minimum read length using --read_min_len 34 (metaphlan -h), in this case all reads with length below 34 will be discarded

1 Like
9

Hi @fbeghini, if I set minimum read length to 34, in that case, do you think the profile output will be reliable? Will MetaPhlAn 3.0 be able to classify the reads perfectly?

Thanks
DC7

10

PRJEB23292 is a project which includes SRAs generated with the SOLiD platform. It is not a good idea converting colorspace reads into basespace, and BowTie2 does not support colorspace input. You should consider analyzing the dataset with a tool able to handle such input.

1 Like
11

Sir, in connection with the sequencing platform, I have a query. Can I analyse WGS sequence data generated by 454 GS FLX platform (PRJNA32089) and Illumina Genome analyser II (PRJEB2054)?

NOTE: I am seeing the fastq files from PRJNA32089 are comparatively smaller (mostly less than 300Mb) and single-ended.

Thanks and regards,
DC7

12

GAII reads are OK for MetaPhlAn, however, since MetaPhlAn is designed to work with short reads, on longer reads such as 454 reads you can try running the internal Bowtie2 using local alignment (--bt2_ps very-sensitive-local) instead of the default end-to-end

1 Like
13

Sir, In that case should I set a --min_alignment_len value other than the default 100? I’m getting a message like this:

Warning! bt2_ps is set to local mode, and min_alignment_len is None, I automatically set min_alignment_len to 100! If you do not like, rerun the command and set min_alignment_len to a specific value.

14

Sir @fbeghini - can you please suggest?

15

I’d suggest to not lower the default value and keep it at 100

1 Like