bgfritz1/Tandem_RNAseq

Costerton Biofilm Center RNA-sequencing analysis pipeline

README

To run the workflow, do the following:

1. Make sure that the necessary prerequisite files are downloaded. At least this should be the human genome fastq and the annotation file

• FASTQ: http://ftp.ensembl.org/pub/release-106/fasta/homo_sapiens/dna/Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz
• Annotation: http://ftp.ensembl.org/pub/release-106/gtf/homo_sapiens/Homo_sapiens.GRCh38.106.chr.gtf.gz

2. Make sure the kraken database is downloaded and accessible. Prebuilt databases can be found here: https://benlangmead.github.io/aws-indexes/k2

3. Edit the metadata.tsv file with your sample information. The metadata file is tab-delimited matrix and should at least have the following column values:

• sample: Where every sample should have a unique sample name
• strategy: either PAIRED or SINGLE referring the whether it was paired-end sequencing or single-end.
• strandedness: refers to the a stranded or directional library prep. 0 = unstranded, 1 = forward-stranded, 2=reverse-stranded.
• Additonal columns are also accepted.

4. Edit the units.tsv file. This contains information about the raw sequence data files. Paths can be either relative or absolute. The unitcolumn refers to samples that were sequenced across multiple lanes on the same flow cell. If the reads are single end, just leave the fq2 column blank.

5. Edit the config/congif.yamlfile with the relevant details. Paths can be relative to the repository directory. Note: Some options in the config file have not yet been implemented.

6. Once everythng is set up, make sure the following programs are available:

• snakemake v6.15.3 I have had problems running the pipeling on snakemake v7
• mamba 0.24.0
• python>3.7
  On computerome, this can be done with:

module load miniconda3/4.12.0
module load mamba-org/mamba/0.24.0
module load snakemake/6.15.3

7. In config/Computerome2/config.yaml change the cluster parameter to the proper Computerome account string.

8. Now, you should be read to go! Navigate to the repository directory (the one containing the config/, workflow/, etc. files) and run the following command: snakemake --profile config/Computerome2. This will start the analysis. It is advisable to run the analysis on a subset of samples first to check that it works as expected.