Fetch architectural stripes¶
The .hdf5 file produced by stripepy call contains various kinds of information, including stripe coordinates, various descriptive statistics, persistence vectors, and more.
While having access to all this information can be useful, usually we are mostly interested in the stripe coordinates, which can be fetched using stripepy view.
# Fetch the first 10 stripes in BEDPE format
user@dev:/tmp$ stripepy view 4DNFI9GMP2J8.10000.hdf5 | head
chr1 910000 960000 chr1 930000 3590000
chr1 1060000 1110000 chr1 1080000 3540000
chr1 1570000 1620000 chr1 1600000 2590000
chr1 1600000 1670000 chr1 880000 1620000
chr1 1670000 1700000 chr1 1680000 2610000
chr1 1730000 1780000 chr1 1750000 2570000
chr1 1780000 1840000 chr1 1780000 2580000
chr1 1890000 1940000 chr1 1920000 3540000
chr1 1940000 2020000 chr1 1960000 3590000
chr1 2020000 2060000 chr1 2020000 3550000
# Redirect stdout to a file
user@dev:/tmp$ stripepy view 4DNFI9GMP2J8.10000.hdf5 > stripes.bedpe
# Compress stripes on the fly before writing to a file
user@dev:/tmp$ stripepy view 4DNFI9GMP2J8.10000.hdf5 | gzip -9 > stripes.bedpe.gz
Output customization and filtering¶
When viewing the stripes, several optional parameters are available to customize the output.
The --relative-change-threshold option allows you to set a cutoff value (defaulting to 5.0) for filtering stripes based on their relative change.
This relative change is calculated as the ratio between the average number of interactions found inside a stripe and the number of interactions in a neighborhood immediately outside of the stripe.
If you are interested in the biodescriptors associated with each individual stripe, you can pass --with-header and --with-biodescriptors when calling stripepy view.
This is the output generated by running stripepy view on the .hdf5 generated using stripepy call v1.1.1.
Files generated by older versions of StripePy may have different columns.
user@dev:/tmp$ stripepy view 4DNFI9GMP2J8.10000.hdf5 --with-biodescriptors --with-header | head
chrom1 start1 end1 chrom2 start2 end2 top_persistence inner_mean inner_std outer_lsum outer_lsize outer_rsum outer_rsize min q1 q2 q3 max outer_lmean outer_rmean outer_mean rel_change
chr1 910000 960000 chr1 930000 3590000 0.3984904019 0.2506571890861574 0.14123131812515843 144.79589039186396 801 192.25135582429806 8010.0 0.17139833204774585 0.22938081658911763 0.28656944403925566 0.9741568863537948 0.18076890186250183 0.24001417705904876 0.2103915394607753 19.138435760573497
chr1 1060000 1110000 chr1 1080000 3540000 0.0826359687 0.23019685453871336 0.14481608064533394 186.18030631678906 741 179.64345985134207 7410.0 0.1539575922232785 0.21018481227951455 0.2710230083036015 0.9903418421799679 0.2512554741117261 0.24243381896267485 0.246844646537200486.744238626207448
chr1 1570000 1620000 chr1 1600000 2590000 0.04103011280000002 0.33195798369580404 0.10697974882795283 99.02697827900961 300 85.58022773213244 300 0.10509240613975727 0.2710230083036015 0.3152772184192718 0.3662448898065007 0.9887477925105556 0.3300899275966987 0.2852674257737748 0.3076786766852368 7.891124361343245
chr1 1600000 1670000 chr1 880000 1620000 0.10798038449999997 0.34673478460468343 0.12547401272240433 79.95811315769556 225 63.18147668278408 225 0.0 0.25904999836303577 0.33447322272887486 0.4155250840484962 0.9887477925105556 0.3553693918119803 0.2808065630345959 0.3180879774232881 9.0059383612837
chr1 1670000 1700000 chr1 1680000 2610000 0.08521339110000004 0.30510000180174507 0.11602295320194354 84.13794539599031 282 71.90225464650885 282 0.0 0.22938081658911763 0.304010183863723 0.37277167877770423 0.8753282776351561 0.29836150849641957 0.2549725342074782 0.2766670213519489 10.276967710447305
chr1 1730000 1780000 chr1 1750000 2570000 0.09549401749999997 0.34157106048803376 0.12939228310023276 66.96694495052422 249 77.44100032822071 249 0.06630592590798857 0.25245019336736707 0.32535592427102433 0.41427461878487365 0.9374989352738993 0.26894355401816955 0.3110080334466695 0.28997579373241955 17.792956471126924
chr1 1780000 1840000 chr1 1780000 2580000 0.14961356020000005 0.31446872398046843 0.14174768874612398 89.65252960337472 243 73.53776985594494 243 0.0 0.2202635181312671 0.28656944403925566 0.3761154144433587 0.9150948504497306 0.3689404510426943 0.3026245673084154 0.33578250917555486 6.347497148501883
chr1 1890000 1940000 chr1 1920000 3540000 0.13643510830000005 0.27087952940479454 0.15589512088714813 98.34422915113818 489 137.9512119037385 489 0.0 0.17139833204774585 0.2453610817780414 0.3592307814635864 0.989227567682685 0.20111294304936234 0.2821088177990563 0.24161088042420928 12.113961477726793
chr1 1940000 2020000 chr1 1960000 3590000 0.05824488140000006 0.267059000791004 0.1518633129658817 138.54936114286124 492 138.81994263073136 492 0.0 0.17139833204774585 0.2453610817780414 0.34858989163711346 0.9751278353396942 0.28160439256679115 0.2821543549405109 0.281879373753651 5.257700400455457
If you are working in Python, you might want to take a look at the classes Result and ResultFile.
Coordinate transformation¶
The --transform option provides control over how stripe coordinates are presented in the output.
By default, no transformation is applied.
However, you can specify transpose_to_ut to transpose coordinates to the upper triangular part of the contact map, or transpose_to_lt to transpose them to the lower triangular part, which can be useful for specific downstream analyses or visualization preferences.