metatranscriptomics_in_DS

Metatranscriptomics study of the Down’s Syndrome model mice fed high-fat diet

This repository contains the code usd to generate the results of the analysis of the metatranscriptomes data. The data were obtained to determine the microbial gene activity in the faecal microbiome samples from Wild Type and Down Syndrome model mice fed high-fat diet.

Partial result files can be found here: http://regulomics.mimuw.edu.pl/~ilona/metatranscriptomics_in_DS/

Quantification of total-RNA sequences

We run the seed-kraken tool with bash commands:

for f in /path_to_fasta_files/*_fwd.fasta; do echo $f output/basename $f.report; kraken --thread $THREADS --fasta-input --preload -db $KRAKEN_DB_NAME --output /output_folder/basename $f.report $f; done

and then:

for f in /output_folder/*report; do echo basename $f; (kraken-report --db $KRAKEN_DB_NAME $f >$f.csv); done

The database for seed-kraken was built based on a collection of sequences from RefSeq (Feb. 2016): Bacteria, Viruses, Fungi, H. sapiens and M. musculus. Sequences needed to have version_status=='latest' and assembly_level = "Complete Genome" or assembly_level = "Chromosome".

Defining differentially expressed genes identified in contigs

R, package DeSeq2 from Bioconductor:

dds <- DESeq2::DESeqDataSetFromMatrix(countData = countData, colData = coldata, design = ~ condition)
dds <- DESeq2::estimateSizeFactors( dds ) ddsp <- DESeq2::estimateDispersions( dds, fitType="parametric" ) ddsp <- DESeq2::nbinomWaldTest( ddsp )
result <- results(dds)

Assessing gene expression levels from raw reads

We run the MetaGeneMark tool with bash commands:

folder_with_MetaGeneMark/mgm/gmhmmp -a -d -m folder_with_MetaGeneMark/mgm/MetaGeneMark_v1.mod input_file.fasta -o output_file_mgm

Then for the predicted genes we found orthologs using eggNOGmapper with the default settings using a command-line, also because files were too big to be processed with the online tool. The command used was:

python folder_with_eggNOGmapper/eggnog-mapper-1.0.3/emapper.py -i meta_genemark_predicted_sequences/input_file.faa --output output_file_eggNOG --cpu 10 -m diamond