#!/bin/bash
#
# Align reads against reference genome.

PROJECT_DIRECTORY="/groups/umcg-griac/tmp01/rawdata/$(whoami)/rnaseq"
# Store genome index in this location:.
GENOME_INDEX="${PROJECT_DIRECTORY}/step3/genome_index"
mkdir -p "${GENOME_INDEX}"
# Store the generated `Aligned.sortedByCoord.out.bam` in this dir.
ALIGNMENT_OUTPUT="${PROJECT_DIRECTORY}/step3/alignment/"
mkdir -p "${ALIGNMENT_OUTPUT}"

# 1) Generate genome index.
#
# N.B.:
# - We're assuming a read size of 100 bp (--sjdbOverhang 100). An alternative
#   cut-off is 150, for low-input methods. In general, refer back to the
#   previous quality control steps if you are unsure about the size. In case of
#   reads of varying length, the ideal value is max(ReadLength)-1.
# - We're using gzip compressed reference data (--readFilesCommand zcat), i.e.,
#   .gtf.gz and fa.gz. If not, you can remove the `zcat` flag.

# Storage location reference data (in this case on Gearshift).
REFERENCE_DATA="/groups/umcg-griac/prm03/rawdata/reference/genome"
GTF_FILE="${REFERENCE_DATA}/Homo_sapiens.GRCh38.100.gtf.gz"
FASTA_FILE="${REFERENCE_DATA}/Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz"

STAR \
    --runThreadN 8 \
    --runMode genomeGenerate \
    --readFilesCommand zcat \
    --sjdbOverhang 100 \
    --genomeFastaFiles ${FASTA_FILE} \
    --sjdbGTFfile ${GTF_FILE} \
    --genomeDir ${GENOME_INDEX}


# 2) Do the actual alignment.
#
# N.B.:
# - We are assuming paired-end, gzip compressed (--readFilesCommand zcat)  FastQ
#   files.

# The compressed, paired-end, FastQ's after trimming (step 2).
R1="${PROJECT_DIRECTORY}/step2/sample1_R1_paired.fastq.gz"
R2="${PROJECT_DIRECTORY}/step2/sample1_R2_paired.fastq.gz"

STAR \
    --runThreadN 8 \
    --readFilesCommand zcat \
    --readFilesIn "${R1}" "${R2}" \
    --outSAMtype BAM SortedByCoordinate \
    --genomeDir ${GENOME_INDEX} \
    --outFileNamePrefix "${ALIGNMENT_OUTPUT}"