CLIP ANALYSIS LEGACY (analyze_motifs)

Runs HOMER and kmer enrichment aspects from clip_analysis_legacy to generate useful motif analyses.

Requirements:

• HOMER
• EMBOSS
• bedtools
• python=2.7
• matplotlib
• seaborn
• numpy
• pandas
• bedtools=2.26
• pybedtools
• htseq
• bx-python
• scikit-learn
• cairosvg=1.0.22

(or run the create_environment.sh script - don't forget to source activate analyze_motifs)

Installation:

python setup.py install

Usage:

analyze_motifs \
--peaks inputs/204_01.basedon_204_01.peaks.l2inputnormnew.bed.compressed.Fc3Pv3.bed \  # BED file (from input normalization or IDR)
--species hg19_v19 \  # sets prefix for finding region bedfiles (see clip_analysis_legacy/data/regions)
--genome_fasta inputs/all.fa \  # fasta file
--k 6 \  # k for kmer enrichment score search
--nrand 1 \  # sets the number of random bedfile regions to generate to form the background
--out_file outputs/204_01_RBFOX2.svg \  # output motif image file
--out_pickle_file outputs/204_01_RBFOX2.pickle \  # output pickle
--out_homer_dir outputs/204_01_RBFOX2_homer_out  # directory by which this script uses to store all outputs (will tar this directory at the end)

(or see examples/run_analyze_motifs.204_01.sh script)

Regarding output parameters...

In the above example, you do NOT need the 204_01_RBFOX2_homer_out/ to exist, but you DO need to make sure outputs/ exists

Notes about installation:

Sometimes compseq will fail if the proper C-libraries aren't found. If this is the case, make sure to append the proper libraries onto your LD_LIBRARY_PATH (on TSCC, the libraries are located in /opt/intel/ (see create_environment.sh).

Notes about data/regions:

This package expects the following files inside data/regions:

• ${SPECIES}_cds.bed
• ${SPECIES}_proxintron500.bed
• ${SPECIES}_distintron500.bed
• ${SPECIES}_five_prime_utrs.bed
• ${SPECIES}_three_prime_utrs.bed
• ${SPECIES}_exons.bed
• ${SPECIES}_genes.bed

where $(SPECIES) is something like: 'hg19' or 'hg19_v19'. Analyze_motifs needs these regions to define where to select the random backgrounds from.

The following species are supported:

• hg19
• hg19_v19
• ce10
• GRCh38
• GRCh38_v24
• mm9
• mm10_vM10

Using nonsupported species:

If you have the proper GTF file, you can use the create_region_bedfiles script inside the annotator package to generate these files, and put them into data/regions if they don't exist there already. You will need to reinstall the package after doing so.

https://github.com/Yeolab/annotator/