Bacteria noncoding transcripts analysis

• This pipeline takes paired metagenomic sequencing (MGX) and metatranscriptomic sequencing (MTX) data as input, and annotates intergenic regions with RNA expression
• Supports paired end / single end data

Preprocess MGX reads

• Trim adaptor with trimgalore
• Remove unwanted sequence, eg. host sequence like human genome for gut metagenome data (optional)
• Run metaphlan for taxonomy profiling

# paired end data
snakemake --snakefile snakefiles/preprocessing.snakefile --configfile config/preprocessing/test-pe.mgx.yaml

# single end data
snakemake --snakefile snakefiles/preprocessing.snakefile --configfile config/preprocessing/test-se.mgx.yaml

Preprocess MTX reads

• Trim adaptor
• Remove unwanted sequence (optional)
• Run metaphlan for taxonomy profiling
• Infer strandness of the RNA library based on marker gene mapping result

# paired end data
snakemake --snakefile snakefiles/preprocessing.snakefile --configfile config/preprocessing/test-pe.mtx.yaml

# single end data
snakemake --snakefile snakefiles/preprocessing.snakefile --configfile config/preprocessing/test-se.mtx.yaml

Assemble MGX reads and annotate contigs

• Assemble MGX reads with megahit
• Run prodigal for gene prediction
• Annotate predicted gene with Pfam & hmmsearch
• Search contig for known noncoding RNA with Rfam & cmsearch (optional)

# paired end data
snakemake --snakefile snakefiles/mgx-analysis.snakefile --configfile config/mgx-analysis/test-pe.yaml

# single end data
snakemake --snakefile snakefiles/mgx-analysis.snakefile --configfile config/mgx-analysis/test-se.yaml

Mapping MTX reads

• Map MTX reads to paired contigs
• Assemble transcripts with stringtie
• Annotate transcripts in a gene centric manner

# paired end data
snakemake --snakefile snakefiles/mtx-analysis.with.mgx.snakefile --configfile config/mtx-analysis-with-mgx/test-pe.yaml

# single end data
snakemake --snakefile snakefiles/mtx-analysis.with.mgx.snakefile --configfile config/mtx-analysis-with-mgx/test-se.yaml

Downstream analysis

• Get transcripts at intergenic regions with distance to nearest CDS >= 16nt
• Retrieve intergenic regions containing these transcripts
• Run FragGeneScan on these intergenic regions to predict candidate CDS
• Run cmsearch on these intergenic regions to annotate known RNA
• We only consider transcripts that does not overlap with known RNAs and coding regions for downstream analysis