Humann3 empty output

Microbial community profiling HUMAnN
1

Hi all,

I am running humann3 for some of my samples and I keep getting empty reads. The samples are from the fish gut and I did RNA sequencing to see how antibiotics and probiotics can change host gene expression. But now I found out that by using Humann3 or Kraken I can get the microbial community as well (while expecting to lose a lot of host reads). So I used Knead data to get rid of the host genes and now while I am using the output from Kneaddata, the humann3 output is empty. Here is a detailed output of my command. I also uploaded my sequences on NCBI SRA and from the NCBI database I can see at least one million reads should be for the microorganism (please see the screenshot or this link (https://trace.ncbi.nlm.nih.gov/Traces/index.html?view=run_browser&acc=SRR21087210&display=analysis). And also I can get results using Kraken.

Screen Shot 2023-05-02 at 1 00 12 PM

05/02/2023 01:11:26 PM - humann.humann - INFO: Running humann v3.6
05/02/2023 01:11:26 PM - humann.humann - INFO: Output files will be written to: /home/jsadeghi/scratch/Javad/Ch4/2_Humann
05/02/2023 01:11:26 PM - humann.humann - INFO: Writing temp files to directory: /home/jsadeghi/scratch/Javad/Ch4/2_Humann/35B_cat_humann_temp
05/02/2023 01:11:26 PM - humann.utilities - INFO: File ( /home/jsadeghi/scratch/Javad/Ch4/1_Knead_data_cleaned/35B_cat ) is of format: fastq
05/02/2023 01:11:26 PM - humann.utilities - DEBUG: Check software, metaphlan, for required version, 3.0
05/02/2023 01:12:09 PM - humann.utilities - INFO: Using metaphlan version 3.0
05/02/2023 01:12:09 PM - humann.utilities - DEBUG: Check software, bowtie2, for required version, 2.2
05/02/2023 01:12:11 PM - humann.utilities - INFO: Using bowtie2 version 2.5
05/02/2023 01:12:11 PM - humann.humann - INFO: Search mode set to uniref90 because a uniref90 translated search database is selected
05/02/2023 01:12:11 PM - humann.utilities - DEBUG: Check software, diamond, for required version, 2.0.15
05/02/2023 01:12:11 PM - humann.utilities - INFO: Using diamond version 2.0.15
05/02/2023 01:12:11 PM - humann.config - INFO:
Run config settings:

DATABASE SETTINGS
nucleotide database folder = /scratch/st-spakpour-1/envs/humann3/installed_databases/chocophlan
protein database folder = /scratch/st-spakpour-1/envs/humann3/installed_databases/uniref
pathways database file 1 = /scratch/st-spakpour-1/envs/humann3/lib/python3.8/site-packages/humann/data/pathways/metacyc_reactions_level4ec_only.uniref.bz2
pathways database file 2 = /scratch/st-spakpour-1/envs/humann3/lib/python3.8/site-packages/humann/data/pathways/metacyc_pathways_structured_filtered_v24
utility mapping database folder = /scratch/st-spakpour-1/envs/humann3/installed_databases/utility_mapping

RUN MODES
resume = False
verbose = False
bypass prescreen = False
bypass nucleotide index = False
bypass nucleotide search = False
bypass translated search = False
translated search = diamond
threads = 1

SEARCH MODE
search mode = uniref90
nucleotide identity threshold = 0.0
translated identity threshold = 80.0

ALIGNMENT SETTINGS
bowtie2 options = --very-sensitive
diamond options = --top 1 --outfmt 6
evalue threshold = 1.0
prescreen threshold = 0.01
translated subject coverage threshold = 50.0
translated query coverage threshold = 90.0
nucleotide subject coverage threshold = 50.0
nucleotide query coverage threshold = 90.0

PATHWAYS SETTINGS
minpath = on
xipe = off
gap fill = on

INPUT AND OUTPUT FORMATS
input file format = fastq
output file format = tsv
output max decimals = 10
remove stratified output = False
remove column description output = False
log level = DEBUG

05/02/2023 01:12:11 PM - humann.store - DEBUG: Initialize Alignments class instance to minimize memory use
05/02/2023 01:12:11 PM - humann.store - DEBUG: Initialize Reads class instance to minimize memory use
05/02/2023 01:12:33 PM - humann.humann - INFO: Load pathways database part 1: /scratch/st-spakpour-1/envs/humann3/lib/python3.8/site-packages/humann/data/pathways/metacyc_reactions_level4ec_only.uniref.bz2
05/02/2023 01:12:33 PM - humann.humann - INFO: Load pathways database part 2: /scratch/st-spakpour-1/envs/humann3/lib/python3.8/site-packages/humann/data/pathways/metacyc_pathways_structured_filtered_v24
05/02/2023 01:12:33 PM - humann.search.prescreen - INFO: Running metaphlan …
05/02/2023 01:12:33 PM - humann.utilities - DEBUG: Using software: /scratch/st-spakpour-1/envs/humann3/bin/metaphlan
05/02/2023 01:12:33 PM - humann.utilities - INFO: Execute command: /scratch/st-spakpour-1/envs/humann3/bin/metaphlan /home/jsadeghi/scratch/Javad/Ch4/1_Knead_data_cleaned/35B_cat -t rel_ab -o /home/jsadeghi/scratch/Javad/Ch4/2_Humann/35B_cat_humann_temp/35B_cat_metaphlan_bugs_list.tsv --input_type fastq --bowtie2out /home/jsadeghi/scratch/Javad/Ch4/2_Humann/35B_cat_humann_temp/35B_cat_metaphlan_bowtie2.txt
05/02/2023 01:16:16 PM - humann.utilities - DEBUG: b’’
05/02/2023 01:16:16 PM - humann.humann - INFO: TIMESTAMP: Completed prescreen : 223 seconds
05/02/2023 01:16:16 PM - humann.search.prescreen - INFO: Total species selected from prescreen: 0
05/02/2023 01:16:17 PM - humann.search.prescreen - DEBUG:

No species were selected from the prescreen.
Because of this the custom ChocoPhlAn database is empty.
This will result in zero species-specific gene families and pathways.

05/02/2023 01:16:17 PM - humann.humann - INFO: TIMESTAMP: Completed custom database creation : 0 seconds
05/02/2023 01:16:17 PM - humann.humann - DEBUG: Custom database is empty
05/02/2023 01:16:17 PM - humann.store - DEBUG: Initialize Reads class instance to minimize memory use
05/02/2023 01:18:01 PM - humann.utilities - DEBUG: Remove file: /home/jsadeghi/scratch/Javad/Ch4/2_Humann/35B_cat_humann_temp/tmp6xxu3ry_/tmp4hjr3d5m
05/02/2023 01:18:01 PM - humann.search.translated - DEBUG: Convert unaligned reads fastq file to fasta
05/02/2023 01:19:19 PM - humann.search.translated - INFO: Running diamond …
05/02/2023 01:19:19 PM - humann.search.translated - INFO: Aligning to reference database: uniref90_201901b_full.dmnd
05/02/2023 01:19:19 PM - humann.utilities - DEBUG: Remove file: /home/jsadeghi/scratch/Javad/Ch4/2_Humann/35B_cat_humann_temp/tmp6xxu3ry_/diamond_m8_hy991q3u
05/02/2023 01:19:19 PM - humann.utilities - DEBUG: Using software: /scratch/st-spakpour-1/envs/humann3/bin/diamond
05/02/2023 01:19:19 PM - humann.utilities - INFO: Execute command: /scratch/st-spakpour-1/envs/humann3/bin/diamond blastx --query /home/jsadeghi/scratch/Javad/Ch4/2_Humann/35B_cat_humann_temp/tmp6xxu3ry_/tmpwzdj83dr --evalue 1.0 --threads 1 --top 1 --outfmt 6 --db /scratch/st-spakpour-1/envs/humann3/installed_databases/uniref/uniref90_201901b_full --out /home/jsadeghi/scratch/Javad/Ch4/2_Humann/35B_cat_humann_temp/tmp6xxu3ry_/diamond_m8_hy991q3u --tmpdir /home/jsadeghi/scratch/Javad/Ch4/2_Humann/35B_cat_humann_temp/tmp6xxu3ry_

2

It looks like your sample is dominated by a eukaryote. It’s possible this species isn’t in the MetaPhlAn 3 database. MetaPhlAn is also tuned for identifying species from DNA reads rather than RNA, though I think that’s the less likely explanation here.

Presumably some of your reads should still be mapping the translated search stage of HUMAnN? The log file cut off before that information, so it’s hard to tell. You can likely improve your mapping by using UniRef50 instead of UniRef90. You could then use the infer_taxonomy helper script to make some guesses about the taxonomic attribution of the unclassified hits from translated search.

1 Like