This repository houses the data components used by the public POTAGE web server (http://crobiad.agwine.adelaide.edu.au/potage).

Intended Audiance

This information is relevant for those seeking to:

1. Add unpublished gene expression data sets to their own POTAGE instance
2. Would like to contribute a published gene expression dataset back to the community of POTAGE users.

Contributing Gene Expression Data Sets to POTAGE

POTAGE currently houses a limited number of publically available gene expression data sets:

1. IWGSC RNA-Seq tissue series under expression/001_iwgsc
2. Meiose data under expression/002_meiose

While we plan to add additional public data sets over time, we welcome contributions from the public via pull requests.

Adding expression data sets

Here we present an example workflow for including additional RNA-Seq datasets in POTAGE. We omit the read QC and alignment steps. We assume the user is able to generate a valid BAM file per data point which is to be displayed in POTAGE. The appropriate reference for aligning the reads when using a splice-aware aligner is MiPS_genes_PlusMinus2kb_allTranscripts__AND__HCS_UNMAPPED_CDS.fasta.gz (TO BE ADDED TO REPO).

Alternatively the user may resort to aligning their reads to transcripts, or even quantify the expression values using an alignment-free approach such as Salmon or kallisto.

We include some of the settings used for aligning RNA-Seq reads from the Meiosis dataset

    STAR \
    --runMode alignReads \
    --readFilesIn ${R1} ${R2} \
    --readFilesCommand pigz -dcp2 \
    --outFileNamePrefix ${OUTBASENAME} \
    --outSAMtype BAM SortedByCoordinate \
    --outFilterMultimapScoreRange 0 \
    --outFilterMultimapNmax 5 \
    --outFilterMismatchNoverLmax 0.00 \
    --outFilterMatchNminOverLread 1.0 \
    --outSJfilterOverhangMin 35 20 20 20 \
    --outSJfilterCountTotalMin 10 3 3 3 \
    --outSJfilterCountUniqueMin 5 1 1 1 \
    --alignEndsType EndToEnd \
    --alignSoftClipAtReferenceEnds No \
    --outSAMstrandField intronMotif \
    --alignIntronMax 10000 \
    --alignMatesGapMax 10000 \
    --outSAMattrRGline ID:${SAMPLE} PL:Illumina PU:${SAMPLE} LB:${SAMPLE} SM:${SAMPLE%_?} || exit 1 

Quantify expression using cufflinks from BAM files generated using STAR

For each of the the (per data point - merge if necessary) BAM files run

    cufflinks -G MIPS_genes_PlusMinus2kb_allTranscripts.gtf data_point.bam \
    -o expression/data_point \
    --no-update-check --max-multiread-fraction 1 --library-type fr-unstranded 

In the case of Chinese Spring Meiosis dataset this could be done as follows:

    for f in merged_bams/*.bam; do s=${f##*/}; 
      cufflinks -G MIPS_genes_PlusMinus2kb_allTranscripts.gtf \
      ${f} expression/${s%.bam}; 
    done

Aggregate per-gene FPKM values from individaul data points into a single table

    for d in expression/* ; do   
      LAST=${d}/sorted_genes.fpkm_tracking
      sort ${d}/genes.fpkm_tracking > ${LAST}
    done
    cut -f7 ${LAST} | tr ':-' '\t\t'  > genes.common
    cp genes.common genes.joined
    for d in expression/*; do
      tmp1=$(mktemp tmp.fpkms.joined.XXXXXXXXXX) 
      paste genes.joined <(cut -f10 ${d}/sorted_genes.fpkm_tracking) > ${tmp1} && mv ${tmp1} genes.joined
    done
    printf 'gene_id\tstart\tstop' > header.txt
    for d in expression/*; do
      s=${d##*/}
      printf "\t${s}" >> header.txt
    done
    cat header.txt <(printf '\n') <(tail -n +2 genes.joined) > meiose.tsv 

Create a simple configuration file describing the data set

For example, for the Chinese Spring Meiosis dataset config.cfg might look as follows:

    #Expression dataset configuration
    ShortName Mei
    LongName  Chinese Spring - Meiosis
    Unit      FPKM
    FileName  meiose.tsv

Make files available to your POTAGE instance and re-load the web-app

POTAGE picks up the expression data sets from a directory specified in potage.cfg, by default this is expression. For an additional dataset to be loaded into POTAGE we create a new sub-directory under expression, for example:

    mkdir expression/002_meiose

We then place the file with expression values (meiose.tsv) and the data set configuration file config.cfg in the newly created sub-directory