derekhoward/he_seq

He et al sequencing

DATA

RNASeq reads

Raw data for for SRA accession # SRP065273 was downloaded from https://sra-explorer.info/ A total of 273 files were downloaded from human (102), chimpanzee (72) and macaque (99) datasets.

Running STAR

Creating a genome index

This step needs to be run only once for all samples using the same genome. These indices can take a large amount of space, so we will dedicate a directory for it on the shared lab space (the arg for --genomeDir is the path where the genome index is stored)

• Fasta files (*.fa) are the genome reference assemblies
• Reference genome annotations are in *.gtf or .gff format. These are recommended to use to aid in better assigning reads to genes (improves alignment process).

STAR --runMode genomeGenerate
--runThreadN 12
--genomeDir /external/rprshnas01/netdata_kcni/stlab/Genomic_references/Ensembl/Human/Release_103/USE_THIS_genomeDir/
--genomeFastaFiles /external/rprshnas01/netdata_kcni/stlab/Genomic_references/Ensembl/Human/Release_103/Raw/Homo_sapiens.GRCh38.dna.primary_assembly.fa \ --sjdbGTFfile /external/rprshnas01/netdata_kcni/stlab/Genomic_references/Ensembl/Human/Release_103/Raw/Homo_sapiens.GRCh38.103.gtf

Alignment process for multiple samples for downstream quantification with RSEM

Test out alignment process for a single sample first. This will give you an idea of the resources required and time to complete a single sample.

Example command used for STAR alignment for downstream RSEM (for single sample)

STAR --genomeDir /external/rprshnas01/netdata_kcni/stlab/Genomic_references/Ensembl/Human/Release_103/USE_THIS_genomeDir/ \
     --sjdbGTFfile /external/rprshnas01/netdata_kcni/stlab/Genomic_references/Ensembl/Human/Release_103/Raw/Homo_sapiens.GRCh38.103.gtf \
     --readFilesIn /genome/scratch/Neuroinformatics/dhoward/he_et_al/SRR2815952_RNA-Seq_of_Human_PFC_section_DS1-Human1-S1.fastq.gz \
     --quantMode TranscriptomeSAM \
     --outSAMtype None \
     --readFilesCommand zcat

To perform alignment for all samples, use the star_pipe.sh script as below

./star_pipe.sh /path/to/fastq_files/ /outputpath/ false /external/rprshnas01/netdata_kcni/stlab/Genomic_references/Ensembl/Human/Release_103/USE_THIS_genomeDir/ /external/rprshnas01/netdata_kcni/stlab/Genomic_references/Ensembl/Human/Release_103/Raw/Homo_sapiens.GRCh38.103.gtf

Running RSEM

Example to run all samples

run RSEM on output results from previous to lab folder ./rsem_pipe.sh /external/rprshnas01/netdata_kcni/stlab/he_human_processed/ /.../STAR_results/coord_bams/ /nethome/kcni/dhoward/rsem_reference/

Creating counts matrix

To aggregate the counts from samples into a single counts matrix we use CLI scripts

module load PYTHON/3.6

python RSEM_counts_matrix.py /path/to/RSEM_results/ ./results/output_path.csv

python ensg_to_genesymbol.py ./results/output_path.csv ENSG_to_gene_name.tsv ./results/output_converted_genesymbols.csv