-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathCHANGES
186 lines (141 loc) · 6.73 KB
/
CHANGES
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
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
1.7.1 Wed Jun 05 2019
Rebrand INSTI to GLPhase
Add documentation
Fix some compilation issues
1.7.0 Thu Dec 10 13:11:08 2015
Cleans up some CUDA code
1.6.8 Thu Nov 19 14:42:58 GMT 2015
Adds support for indexed BCF/VCF reading of input haplotypes
Adds genotype calling accuracy test based on 1000GP3 GLs
1.6.6 Wed Jun 03 12:07:14 BST 2015
Added subsetting by region
Added subsetting by sample list
Moved to boost::program_options
Added dropping of extra sites (past 1024)
1.6.2 Fri May 08 15:38:42 BST 2015
Adds more tests for multiallelic sites
Genetic distances are calculated correctly from genetic map
Adds genetic distance inflation option -R
1.6.1 Fri May 01 10:21:10 BST 2015
Accepts input GLs in BCF format
Accepts input GLs with multiple sites at the same position
1.4.13 Tue Nov 25 14:40:51 2014 +0000
Added makefile
Added genetic map to replace ad-hoc estimation of recombination
rate using Waterson's estimator
1.4.12 Tue Jul 8 21:23:43 2014 +0100
Added more filtering options for clustering
1.4.11 Tue Jul 08 21:17:06 BST 2014
glVQ: increased training set size to entire set
added more omp parallelization
fixed bug in logging
fixed bug with -T option where tract length was not used before
logging now also logs sampler state
added support for reclustering
introduced threads option
fixed bug in tract length distance metric
added support for alpha-numeric version "numbers"
1.4.5 Sun May 25 20:26:47 BST 2014
GLs are now compressed using vector quantization (Lloyd's
algorithm on a subset of GLs)
Compressed vectors are now stored in uint32_t instead of char
as the char gets stored in a 32 bit register anyway, duh.
1.4.4 Wed May 21 11:08:34 BST 2014
hmm_work() is now executed in parallel on the host
Input haplotypes for scaffold are now read from tabhaps.gz file
1.4.2 Fri May 16 14:36:24 BST 2014
GLs are now only copied over once if the stride is equal to
the number of samples that GLs exist for
1.4.1 Fri May 16 10:58:52 BST 2014
Moved HMMLike object creation outside of cudaSolve
1.4.0 Wed May 14 17:21:44 BST 2014
Completed work on writing a first GPU based version of hmm_like
all tests pass
introduced NCUDA variable to makefile
1.3.6 Sat Mar 22 11:59:13 GMT 2014
Pulled in all changes from 1.3.4
Fixed bug in main.cpp where -M was not auto-set to start after
simulated annealing burnin generations
1.3.5 Sat Mar 22 11:15:05 GMT 2014
Restructured sampling hierarchy need to find bugs and merge
with version 1.3.4
1.3.4 Fri Mar 21 13:18:10 GMT 2014
Fixed bug in tract length calculation
Haplotypes.hpp now uses bitsets under the hood
1.3.3 Wed Mar 19 15:53:28 GMT 2014
Moved all enum classes to be defined in their respective class
headers
Added option to use tract length instead of hamming distance
for kNN clustering
1.3.2 Sat Mar 15 17:50:00 GMT 2014
Moved selection of graph types to enum class
1.3.1 Mon Mar 10 11:36:56 GMT 2014
Fixed some bugs in the delayed acceptance of the MH class.
1.3.0
Added delayed acceptance MH.
1.2.18 Wed Mar 05 15:05:51 GMT 2014
Scaffold haps reading now checks only chrom and positon before
splitting the whole string.
snp.hpp now also keeps track of chromosome.
1.2.17 Thu Feb 20 13:37:49 GMT 2014
Fixed bug where APP field was being terminated with "," by mistake
1.2.16 Sat Feb 15 22:50:02 GMT 2014
Added Unit tests for initializing sites to scaffold if
scaffold is available.
Fixed bug where only scaffold sites equal to 1 were set.
Removed "using namespace std" from olivier/utils.hpp ...
1.2.15 Tue Feb 11 15:07:31 GMT 2014
VCF output is now in GT:GP:APP format
1.2.14 Mon Jan 27 17:23:53 GMT 2014
First go at setting sites to scaffold if scaffold is available
still need to implement unit tests for this
1.2.13 Mon Jan 27 16:34:12 GMT 2014
Accepting empty scaffolds was a bad idea. Now exiting in this
case again
1.2.11 Mon Jan 27 14:11:35 GMT 2014
OpenHaps now accepts scaffolds that don't contain any sites
1.2.8 Fri Jan 17 15:47:24 GMT 2014
OpenHaps now accepts scaffolds that only contain a subset of
the GL sites.
1.2.7 Thu Jan 16 14:20:38 GMT 2014
Nearest neighbors are now calculated for each haplotype
separately and then half the number of target haplotypes are
kept from each list
1.2.6 Tue Jan 14 18:11:49 GMT 2014
Common site haplotypes are now stored in memory
1.2.5 Tue Jan 14 17:07:50 GMT 2014
Fixed bug where number of haps was passed to relationship.cpp instead of
number of sites
1.2.4 Tue Dec 10 12:29:49 GMT 2013
Added filtering and sorting of hapsSamp samples files
1.2.3 Mon Dec 09 15:32:32 GMT 2013
Added filtering of haps files at input to speed things up
1.2.2 Mon Dec 09 14:01:57 GMT 2013
Added more tests to make sure filtering of sites for scaffold works correctly
1.2.1 Fri Nov 22 16:47:30 GMT 2013
Changed input of scaffold to haps sample format
1.2.0 Thu Nov 21 17:00:46 GMT 2013
Added nearest neighbor clustering a la IMPUTEv2 only using a scaffold
1.1.24 Wed Oct 16 15:23:05 BST 2013
Found another bug related to previous bug
Changed -O1 back to -O3, as the optimization was not a problem after all
1.1.23 Wed Oct 16 15:00:10 BST 2013
Fixed bug in RelationshipGraph::sampleHap() where for graph type 2 only half
the available haplotypes were being sampled from.
Changed -O3 to -O1 for default optimization, as the higher -O's seem to be causing
bugs in my code at RelationshipGraph::init()
1.1.22 Tue Oct 15 15:34:13 BST 2013
1) glphase now prints '999' instead of 'inf' in GP field of VCF output
2) glphase now checks to make sure probabilities are between 0 and 1 (inclusive)
1.1.21 Wed Oct 09 18:33:53 BST 2013
caught bug where main does not exit with error of file could not be loaded
1.1.20 Mon Sep 30 17:36:54 BST 2013
Output VCFs now use GP instead of non-standard AP field
1.1.19 Sun Sep 29 13:34:26 BST 2013
fixed bug where default number of burnin iterations was 46 instead of 56
1.1.18 Sun Sep 29 12:25:48 BST 2013
haps are now initialized according to GLs with flat prior
1.1.1 Thu Sep 12 17:16:03 BST 2013
Added support for clustering of haplotypes
1.0.11 Mon Sep 2 12:23:59 2013
Version used for GoCD