mpdoane2 / fameous-metagenomic-workflow

short read annotation for taxonomic and gene functions

For the current implementation of the FAME metagenomic workflow go to: https://github.com/linsalrob/atavide_lite/blob/main/slurm/README.md

Outputs files from the above implementation:

1. fastq/: Demultiplexed raw fastq files from the sequencing facility
2. fastq_fastp/: These are the tag cleaned files
3. fasta/ fasta: Files converted from Fastq_fastp/ fastq files for mmseqs
4. mmseq/: Results of mmseqs using the UniRef50 database. These are the raw outputs from mmseqs for each individual sample. These files include both the taxonomimc and gene function annotations. If you wish to link the taxonomic and gene function annotations (ie read # X corresondes with taxa A and gene function R), the linking information is in <sample_name>_tophit_report_subsystems.gz.
5. Taxonomy/: parsed results from mmseqs output converting the Kraken2 output into a table format of taxonomic hierarchical categorisation by samples. Number indicate counts of reads assigned to each taxa.
6. subsystems/: parsed results from mmseqs converting the subsystem assignments into a table format with subsystem hierarchies by samples. Numbers indicate counds of reads assigned to each subsystem
   • three sets of outputs are included: A) <subsystem_level>_norm_all.tsv: values are normalized by total number of reads in library B) <subsystem_level>_norm_ss.tsv: values are normalized by the total number of reads with a hit to a subystem C) <subsystem_level>_raw.tsv: values are raw counts
   • It is unclear which way is correct -- try to analyse and see if the results change.
7. megahit/: These are the constructed contigs from the samples
8. vamb/: The results of binning the contigs can be found in this directory.

FAMEous-metagenomic-workflow

This is a step-by-step workflow used to analyze metagenomes. For paired end reads only

The commands are saved as SLURM scripts to run on a local cluster - alter the SLURM parameters for your own HPC

Input data

Place the reads in the fastq directory, copy them over to this directory.

Currently this directory has two test metagenomes

Pre-requisite software

Download Miniconda2 - https://docs.conda.io/en/latest/miniconda.html

Quality control

• If you wish to just run quality control to remove the low quality reads, go to Prinseq directory for more instructions
• For human DNA or other host removal steps followed by prinseq - go to Trimnami directory for more instructions

Single read annotation

• Go to MMseqs2 directory for taxonomic and functional based annotation
• Go to kraken directory for taxa based annotations
• Go to superfocus directory for function based annotations

Assembly

• Go to megahit directory for assembly
• Go to spades directory for assembly

Binning

• Go to metacoag directory
• Go to vamb directory