Comments / Criticisms / Feature requests: https://www.biostars.org/p/157346/

#peak-tool-multi Tool to annotate human hg19 or mouse mm10 aligned ChIP-Seq peak files. This tool also takes multiple ChIP-Seq peak files from different experiments and finds neighbors of the primary peak file and annotate it.

20GB RAM required on Linux

16GB RAM required on MAC OS X >= 10.9

This is a tool in c++ to parse the gencode annotation database file and list(s) of ChIP-Seq peaks in bed file format. Report detailed promoter / gene-body / intergenic / enhancer occupancy in human or mouse. Multi peak option reports neighboring peaks in order to elicit co-acting transcription factors. For e.g. With this feature you can correlate your peaks with ENCODE ChIP-Seq data or a related ChIP-Seq data.

Compiling::

gunzip gencode.v19.annotation.gtf.gz

gunzip enhancers.bed.gz

make

Running::

Example1: ./peak_tool_multi ./test.bed > test.genes.txt

Example Input in BED format

chr1 1778750 1778751 MACS_peak_51 102.12

chr1 1933483 1933484 MACS_peak_57 93.87

chr1 3446145 3446146 MACS_peak_91 85.75

chr1 4003155 4003156 MACS_peak_104 58.09

chr1 5787471 5787472 MACS_peak_121 1325.16

chr1 6473142 6473143 MACS_peak_138 988.16

chr1 7259083 7259084 MACS_peak_154 60.32

chr1 8031408 8031409 MACS_peak_181 750.45

chr1 8319346 8319347 MACS_peak_195 3100.00

chr1 8360245 8360246 MACS_peak_198 261.50

The output is in the following format

chrnum peak_mid peak_name peak_score peak_location gene_name strand isoform isoform_coding_type tss dist_tss

chr1 1778750 MACS_peak_51 102.12 INTRON GNB1 - GNB1-001 protein_coding 1822495 43745

chr1 1933483 MACS_peak_57 93.87 INTRON C1orf222 - C1orf222-007 retained_intron 1935276 1793

chr1 3446145 MACS_peak_91 85.75 INTRON MEGF6 - MEGF6-001 protein_coding 3448012 1867

chr1 4003155 MACS_peak_104 58.09 ENHANCER . . . . . .

chr1 5787471 MACS_peak_121 1325.16 INTERGENIC . . . . . .

chr1 6473142 MACS_peak_138 988.16 EXON HES2 - HES2-002 protein_coding 6484730 11588

chr1 7259083 MACS_peak_154 60.32 INTRON CAMTA1 + CAMTA1-001 protein_coding 6845384 413699

chr1 8031408 MACS_peak_181 750.45 EXON PARK7 + PARK7-004 protein_coding 8014351 17057

chr1 8319346 MACS_peak_195 3100 ENHANCER . . . . . .

Example2: ./peak_tool_multi EXP1.bed EXP2.bed EXP3.bed EXP4.bed > EXP1_EXP2_EXP3_EXP4.genes.txt

Output of multi-peak feature

chr1 714304 MACS_peak_2 158.84 INTERGENIC . . . . . . 714017 MACS_peak_1 68.78 -287 . 714039 MACS_peak_1 170.93 -265 . 714023 MACS_peak_1 245.28 -281 .

chr1 769360 MACS_peak_8 421.42 INTERGENIC . . . . . . 769283 MACS_peak_3 55.43 -77 . 769273 MACS_peak_5 154.26 -87 . 769292 MACS_peak_3 71.24 -68 .

chr1 840155 MACS_peak_10 96.66 INTERGENIC . . . . . . 840097 MACS_peak_6 165.5 -58 . 840026 MACS_peak_11 57.72 -129 . 840075 MACS_peak_8 134.82 -80 .

chr1 840738 MACS_peak_11 137.93 ENHANCER . . . . . . 840097 MACS_peak_6 165.5 -641 . 840026 MACS_peak_11 57.72 -712 . 840075 MACS_peak_8 134.82 -663 .

chr1 911680 MACS_peak_15 241.99 PROMOTER C1orf170 - C1orf170-002 retained_intron 912021 341 911740 MACS_peak_20 271.3 60 281 911707 MACS_peak_31 285.65 27 314 911709 MACS_peak_23 251.1 29 312

chr1 994706 MACS_peak_18 95.46 ENHANCER . . . . . . 994613 MACS_peak_35 66.76 -93 . 995290 MACS_peak_52 57.49 584 . 994655 MACS_peak_37 61.84 -51 .

chr1 1003263 MACS_peak_19 71.56 INTERGENIC . . . . . . 1003034 MACS_peak_37 252.84 -229 . 1003088 MACS_peak_54 281.96 -175 . 1003078 MACS_peak_39 643.19 -185 .

chr1 1003982 MACS_peak_20 104.28 ENHANCER . . . . . . 1003034 MACS_peak_37 252.84 -948 . 1003088 MACS_peak_54 281.96 -894 . 1003078 MACS_peak_39 643.19 -904 .

chr1 1098334 MACS_peak_23 149.14 INTERGENIC . . . . . . 1098987 MACS_peak_50 185.58 653 . 1098482 MACS_peak_70 130.48 148 . 1098535 MACS_peak_51 138.52 201 .

Notes::

a) Make sure you have g++ / clang installed on your system

b) Tested with MACS bed file (Peak summit is preferable)

c) Simplistic assumption of Promoter regions (-2kb - +0.5kb) (can be changed in the program)

d) To check multiple peak files (1k region, can be changed in the program) and elicit neighboring peaks use peak_tool_multi in the multi-ChIP directory.

e) Works perfectly on a Macbook Pro Retina laptop with 16GB RAM and MAC OS X > (10.9.x). The OS efficiently compresses the indexes data-structure in memory.

Linux specific notes

f) Use ZRAM on a Linux system with 16GB RAM or use a system with >= 20 GB RAM to avoid swapping and unnecessary frustrations http://askubuntu.com/questions/174579/how-do-i-use-zram

g) If your Linux system has < 20GB RAM please make sure you have ~30%-50% of your RAM allocated as swap with ZRAM

edit the following zram config file in Ubuntu Linux vim /etc/init/zram-config.conf > mem=$(((totalmem * 30/100 / ${NRDEVICES}) * 1024))