Releases: YaqiangCao/cLoops
Releases · YaqiangCao/cLoops
cLoops v0.93a
Just small updates like the first command line in example/run.sh to keep updates.
cLoops v0.93
- updated DBSCAN algorithm as CHEN Zhaoxiong's version, much faster for clustering, nearly 1/2 time saved compared to old DBSCAN version.
- updated method for visiting PETs, now using CHEN Zhaoxiong's method of binary search, much faster.
- new DBSCAN algorithm designed by us called blockDBSCAN is uploaded and under intense fine-tune for cLoops.
Happy loop-calling.
cLoops v0.92
Some small updates as following:
- Optimized combining clustering results, much faster
- -max_cut option added to use the maximal estimated self-ligation and inter-ligation distance cutoff when multiple eps or minPts feed to cLoops. This option works well for Trac-looping data.
cLoops v0.91
- Initial version of callStripes added, which can call stripes (defined in The Energetics and Physiological Impact of Cohesin Extrusion ). We tested it using K562 H3K27ac HiChIP data, not fully tested in deep-sequenced Hi-C data.
cLoops v0.9
- Pre-set parameters for our tested ChIA-PET, Hi-C and HiChIP data, with the parameter -m (--mode) to set, -m 0 accepts user settings; -m 1 for sharp peak like ChIA-PET data; -m 2 for broad peak like ChIA-PET data; -m 3 for deep sequenced Hi-C (~200 million cis PETs); -m 4 for HiChIP (~100 million cis PETs);
- Multiple minPts options added;
- Fixed bug of merging overlapped loops;
- The generation of self-ligation and inter-ligation PETs distance distribution plot can be determine by user using -plot option;
- Turn off deLoops for further development;
- Roll back cDBSCAN for Yaqiang Cao's version, Zhaoxiong Chen's version is under further improvement;
- fixed several small bugs
- for old option like -j 1, now set -j is OK.
cLoops v0.6
- New cDBSCAN by CHEN Zhaoxiong, much faster for HiChIP and Hi-C data.
- New query method for genomic regions by CAO Yaqiang, much faster and less memory usage for second step loops' significance estimation.
cLoops v0.5
This is version for our original paper.