-
Notifications
You must be signed in to change notification settings - Fork 20
/
Release_Notes.txt
164 lines (109 loc) · 10.4 KB
/
Release_Notes.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
Release Notes for FitHiChIP
---------------------------
Release - Version 12.0 (To do)
---------------------------
1. Differential analysis of HiChIP loops is incorporated with an additional parameter - *BackgroundFDRThr*. This parameter denotes the FDR threshold used for determining the background FitHiChIP loops in the edgeR model. By default the threshold is 1, means every FitHiChIP contact is used as the background. A value of 0.01 would indicate that the FitHiChIP loops with q-value < 0.01 in at least one sample would be used as the background. User can test with different thresholds to vary the background set of loops for edgeR.
Release - Version 11.0 (December 2022)
---------------------------
1. FitHiChIP now support HiChIP interactions in .hic and .cool / .mcool formats, in addition to the earlier formats.
2. Updated configuration files corresponding to these new input options.
3. Updated Docker and Singularity packages.
4. Differential HiChIP loop calling does not require ChIP-seq alignment files as a mandatory option. If users do not have any ChIP-seq alignment file, they can just proceed with the differential analysis without considering the difference in 1D.
5. FitHiChIP output loops are now converted to files compatible with WashU, UCSC and IGV epigenome browsers.
Release - Version 10.0 (April 2022)
---------------------------
1. HiC-pro installation directory is now checked only if user provides HiC-pro valid pairs output. If user provides matrix and bin interval files (obtained from another source), he/she does not require to install HiCPro.
2. HiC-pro version checking is stopped. Although users are requested to use the latest version of HiCPro.
3. Parallel processing is updated. Instead of using mclapply, we use lapply. Some users experienced halt / crash of parallel processing routine.
Major release - version 9.1 (October, 2021)
-----------------------------------------
1. Upgraded merge filtering routine to support Python3, instead of deprecated python2
2. Added support for circular genome. In such a case, the genomic distance between interacting fragments are computed using the minimum of their linear distance and their circular genome specific distance. The configuration file now has a parameter "CircularGenome" which, if 1, denotes that the reference genome is circular. **Note** User must provide the appropriate chromosome size file in the configuration options.
3. Updated the genomic distance based filtering of interactions according to the circular genome.
4. The FitHiChIP output file (*fithic*.bed) has now one extra field, namely "Dist" (20th column) which explicitly mentions the genomic distance between the interacting fragments. Useful for the circular genome.
5. Updated the HiChIP peak calling routine as well. Previously, all the reads from DE, SC, RE and valid pairs (from HiC-pro output) were required as the input set of reads. Now we've relaxed such constraints. If the user does not provide any DE, RE, or SC reads, but only provides the valid pairs, HiChIP peaks will be estimated from the valid pairs itself.
6. Updated README and manual
Release 9.0 (January 10, 2021)
------------------------------
1. Updated singularity, Docker installation, incorporating the latest code and dependencies.
2. Incorporated simultaneous generation of WashU browser compatible tracks for differential analysis.
3. Minor warning fix regarding 0 size of input peaks for a given chromosome.
4. Updated documentation.
Release 8.0 (January 30, 2020)
------------------------------
1. Minor bug fix in differential analysis module
2. Using ggplot2 for plotting spline fit and regression coefficients, in the significant interaction module. Useful for running in computational cluster environment.
Release 7.1 (October 24, 2019)
---------------------------------
Major changes:
1. Updated differential analysis module - support processing ChIP-seq coverage files in BEDGraph format (in addition to process ChIP-seq alignment .bam files). Improved code with much lower running time. Improved documentation.
2. Thoroughly tested support of HiC data. User now can provide HiC contact matrices in either HiC-pro based validpairs / matrix format, or simple bed format (peak file is not required) and specify ALL-to-ALL (4) interaction type to obtain HiC data specific significant interactions.
3. Included code for simulating HiChIP data from input HiC and ChIP-seq (as published in our Nature Communication Paper). Note that this implementation is far from optimal, and was mainly to show the robustness of FitHiChIP. The simulation needs to be much improved.
3. Lower running time in finding statistically significant interactions, by faster data reading and processing.
4. Updated nextflow and docker installation steps.
Minor changes:
1. Added: differential analysis using gzipped input files.
2. Removed: dependency of specifying HiC pro installation directory in the configuration file.
2. Bug fix: differential loops using input files with a subset of chromosomes (even one chromosome).
3. Bug fix: parsing input parameters - error in checking q-value range (invalid arithmetic operator)
Release 7.0 (29th April 2019) (Major release)
-----------------------------------------------
1. Included support for processing pre-computed set of locus pairs along with their observed contact count. The file should have the following seven fields in tab seperated format: chr1,start1,end1,chr2,start2,end2,contact
Such a file can be computed using any HiC or HiChIP data processing pipeline.
In the sample configuration files, parameter (A.4) corresponds to this new parameter.
2. Completely updated methodology of differential HiChIP loop finding using FitHiChIP loops of two categories, each with one or more replicates.
The script "Differetial_Analysis_Script.sh" contains the updated commands.
We have also included test data of differential loop finding under the folder "TestData/DiffLoopData"
3. Modified HiChIP coverage computation for different types of inputs (i.e. either valid pairs file, or matrix + bin files, or pre-computed locus pairs file)
4. Removed redundant and optional parameters from the command line options, so that users should enter only the essential parameters for executing FitHiChIP. Sample configuration files are updated.
Release 6.0 - January 17, 2019 (second draft of BiorXiv)
--------------------------
1. Supports ChIP-seq peak files with or without arbitrary header lines.
2. Minor update in plotting spline fit of regression coefficients (loop calling module).
3. Updated code to convert the significant loops into WashU epigenome browser compatible format. A gzipped file "*_WashU.bed.gz" along with its tabix indexed file are created which can be directly used in WashU genome browser for loop visualization.
4. Included routines to check the dependency packages and their installation in the host system, and report users accordingly.
5. Updated configuration files so that users can readily run the test data provided with this package.
Release 5.1 - Sept 25, 2018
--------------------------
1. Creation of a summary .html file in the specified output directory. This summary file lists all the important output files the user should check, including the significant interactions. It also plots distribution of interactions according to the genomic distance, FDR thresholds, etc.
2. Support for gzipped file in the validpairs input, matrix input (from HiC-pro) and also for the input peak file. It'll enable users to save storage space.
3. Minor fix in the parsing configuration file routine.
4. Minor modification in the spline and interaction plots.
Release 5.0 - Sept 21, 2018 (corresponding to first draft in BiorXiv)
--------------------------
1. Introduced differential HiChIP loop finding routine. Input: two categories of FitHiChIP loops, each with multiple replicates.
2. Included a separate script for peak calling from the HiChIP data itself. The generated peaks can be used for significant loop calling.
3. Included various utility scripts:
3.1: producing contact matrix of varying resolutions (bin size)
3.2: merging ChIP-seq alignments (for producing ChIP peaks from the merged alignment)
3.3: generating ChIP-seq coverage from a ChIP alignment file. Useful for finding replicate specific CHIP coverage to be applied to the differential analysis module.
4. Minor bug fix in merged filtering code.
5. Added plotting of distance vs contact count for the significant loops.
6. Cleaned code for discarding various redundant / experimental parameters (internal).
Release 4.1 - April 3, 2018
--------------------------
1. Reference files such as fasta sequence (corresponding to the reference genome), mappability file, restriction fragment file, are made as optional parameters. User can proceed without specifying these parameters (keeping the entries blank in the configuration file). The output interaction file is of the same format as before, except that, without any such genome specific input files, those fields have 0 values.
2. Included support for gzipped Valid pairs and peak files.
3. Conditional plotting of mappability, GC content, and RE sites related features, based on the inpuyt settings.
Release 4.0.1 - March 24, 2018 (major release)
--------------------------
1. Bias regression technique is fully updated.
2. Distance decay modeling and spline fitting (equal occupancy binning) is also fully modified.
3. Bias regression using both peak and non-peak specific coverage bias, as well as ICE bias (implemented in HiC-pro pipeline) are supported.
4. Introduces stringent background (using only peak to peak locus pairs) for significant loop calling for peak-to-all foreground. Corresponding version of FitHiChIP is termed as FitHiChIP(S). Earlier implemented peak-to-all locus pair specific background modeling is now referred to FitHiChIP(L).
5. Introduces "Overwrite" parameter in the configuration file, which if 1, overwrites existing results.
Release 3.0 - Nov 28, 2017
--------------------------
1. Included a configuration file as input, where user would specify the input files as well as execution parameters.
2. Update of merged filtering code.
3. Adjustment of code corresponding to bedtools version 2.26.0
Release 2.0 - Nov 15, 2017
--------------------------
1. Updated bias regression.
2. Peak and non-peak specific coverage bias definition.
3. First draft of merged filtering (of adjacent loops) routine.
4. Minor bug fix regarding distance thresholding of loops.
Release 1.0 - Oct 19, 2017
--------------------------
1. First release of FitHiChIP.
2. Draft version of bias regression.