forked from al2na/methylKit
-
Notifications
You must be signed in to change notification settings - Fork 0
/
NEWS
230 lines (129 loc) · 8.99 KB
/
NEWS
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
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
methylKit 0.9.2
--------------
IMPROVEMENTS AND BUG FIXES
* A bug introduced with 0.9.1 is fixed. The bug occured when unite() function
is used with destrand=TRUE argument. It returned less number of CpGs then it was
supposed to, although returned CpGs had correct methylation and coverage values.
methylKit 0.9.1
--------------
NEW FUNCTIONS AND FEATURES
* objects takes less memory and they are reorganized. updateMethObject() updates
the objects from previous versions the latest version.
* New batch effect control functions are implemented. You can control if certain principal
components are associated with batch effects and remove those components from your
data. see ?assocComp amd ?removeComp. In addition, if you have corrected for the
batch effects via other methods, you can reconstruct a corrected methylBase object (
see ?reconstruct). Check the "batch effects" section in the vignette.
IMPROVEMENTS AND BUG FIXES
* unite() function takes less time due to use of data.table::merge
* fixes a bug appeared with R 3.02, where getting exon and intron coordinates
from BED12 files produced an error.
* tested with R 3.2 and matching bioconductor packages
* data(methylKit) has the new objects
methylKit 0.5.7
--------------
IMPROVEMENTS AND BUG FIXES
* tested with R 3.0 and matching bioconductor packages
* deprecated matchMatrix import from IRanges was causing a problem with package installation. v0.5.7 fixes that problem.
* Now there is no need for "chr" string in BED files when reading them as annotation files. Some assemblies do not have the "chr"" string in their chromosome names.
methylKit 0.5.6
--------------
NEW FUNCTIONS AND FEATURES
* new arguments for clusterSamples() and PCASamples() functions are added. With
the new options "sd.filter","sd.threshold","filterByQuantile" are added. These
options help finetune how low variation bases/regions are discarded prior to
clustering or PCA. See ?PCASamples and ?clusterSamples() for details on the
new options
* FAQ section added to the vignette
* show methods added for each class. Now, typing the variable name containing the
object will display concise information about the contents of the object.
* Subsetting objects via "[" notation is now enabled. You can subset rows of the
objects and it will return a new object rather than just a data frame.
IMPROVEMENTS AND BUG FIXES
* tileMethylCounts() error is fixed. Error occurred when tilling sparsely
covered small chromosomes like chrM in human RRBS data.
* read.bismark() can deal with Bismark output with Bowtie2. Bowtie2 can put gaps
in the alignment, now read.bismark() can deal with those gaps when parsing the
SAM file.
* Coverage columns are coerced to integer when reading generic methylation per
base files. BSMAP scripts can produce a methylation ratio file where coverages
(or "effective CT counts" as they are called) are not always integers, which
causes a problem in the downstream analysis. Now, these non-integer columns
are rounded to nearest integer while reading.
See http://zvfak.blogspot.com/2012/10/how-to-read-bsmap-methylation-ratio.html
for example usage of this functionality.
methylKit 0.5.5
--------------
IMPROVEMENTS AND BUG FIXES
* Differential methylation percentage calculation bug fixed. The bug occurred when "min.per.group" argument used in unite() function and when "weighted.mean=TRUE" in calculateDiffMeth() function.
* plotTargetAnnotation bug is fixed. Bug occured when "precedence" argument set to FALSE.
methylKit 0.5.4
--------------
IMPROVEMENTS AND BUG FIXES
* Examples added to help pages
* Changes to DESCRIPTION for complying with the bioconductor guidelines
* unused "cor" option removed from PCASamples() function
* some irrelevant functions are not exported (i.e they are not public) anymore.
* getContext() looks for the correct slot name now
methylKit 0.5.3
--------------
NEW FUNCTIONS AND FEATURES
* new function adjust.methylC() can be used to adjust measured 5mC levels by measured 5hmC levels
IMPROVEMENTS AND BUG FIXES
* pool() function bug fixed where one of the groups have one sample the pool function was not returning correct values
* calculateMethDiff() function option "SLIM" is now working. If set to TRUE SLIM method for q-value calculation will be used. If FALSE, p.adjust with method="BH" option will be used for P-value correction.
* read.bismark() function now works correctly under Windows.
* read.bismark() bug occurring when the reads are paired and overlapping is fixed now.
* read.bismark() bug occurring when the alignment is done in non-directional manner is fixed now. However, illumina sequencing protocols are directional and you are unlikely to have encountered this error if you aligned your sequences in a directional manner.
* getCorrelation() function has a new option called "method" can take the value of "spearman","person" or "kendall"
methylKit 0.5.2
--------------
NEW FUNCTIONS AND FEATURES
* new function pool() sums up coverage, numCs and numTs values within each group so one representative sample for each group will be created in a new methylBase object.
* new function normalizeCoverage() normalizes coverage and associated number of Cs and number of Ts values between samples using a scaling factor derived from the ratio between mean or median of coverage distributions of samples.
methylKit 0.5.1
--------------
IMPROVEMENTS AND BUG FIXES
* calculateDiffMeth() can now deal with differential methylation calculations where one group has multiple samples but the other one has only one.
methylKit 0.5
--------------
NEW FUNCTIONS AND FEATURES
* new function reorganize() can be used to reorganize methylRawList and methylBase objects to create new objects from subset of samples or can be used to change the order of samples and treatment vector.
* new function bedgraph() can output UCSC bedgraph files for methylRaw and methylDiff objects
* new function percMethylation() extracts percent methylation values from a methylBase object and returns a matrix
* unite() now can merge bases/regions that are not covered by all samples by setting "min.per.group" option
* PCASamples() has new options "transpose" and "sd.threshold". Now, one can do PCA analysis on the transposed % methylation values. "sd.threshold" is for removing rows with small variation prior to PCA analysis .see ?PCASamples for details
IMPROVEMENTS AND BUG FIXES
* regionCounts() bug that appeared with data.table 1.8.0 is fixed
* unite() bug that appeared with R 2.15 is fixed
* calculateDiffMeth() can deal with NA values introduced by setting "min.per.replicate" option of unite() function
* PCASamples() now uses "prcomp" function to do the PCA analysis
* external annotation data (cpgi.hg18.bed.txt and refseq.hg18.bed.txt) that comes with the package is now only a subset of full datasets. Do not use them for your own analysis, download the complete annotation data from UCSC or other sources in BED format.
methylKit 0.4.1
--------------
NEW FUNCTIONS AND FEATURES
* new function select() can be used to subset methylRaw, methylBase and methylDiff objects to create new objects with a subset of methylation data useful if you want to use only a particular portion of methylation events on the genome.
IMPROVEMENTS AND BUG FIXES
* read.bismark bug fixed, now it should be able to save methylation call files with no problem
methylKit 0.4
--------------
NEW FUNCTIONS AND FEATURES
* New read.bismark() function can read directly from a sorted SAM file output by Bismark aligner, the function can save methylation calls as text files or read them as methylRaw object to be used in analysis
* calculateDiffMeth() can do all differential methylation calculations using multiple-cores if 'num.cores' is set to an integer denoting how many cores should be used.
* getCoverageStats() & getMethylationStats() have a new option called 'labels', if set to FALSE, no labels will be drawn on top of the histogram bars.
IMPROVEMENTS AND BUG FIXES
* cov.bases option in tileMethylCounts() now works
* methylRaw,methylBase and methylDiff objects have a new slot 'resolution', which designates
whether methylation information is base-pair resolution or regional resolution. allowed values 'base' or 'region'
* getCoverageStats() & getMethylationStats() now print the sample ids and methylation context automatically in the
title when plotting
* getCoverageStats() & getMethylationStats() takes extra options to be passed to hist() function
* destrand option of unite() function will be over-ridden when methylRawList to be united contains regions rather than bases
* unite() function checks if supplied elements of methylRawList object have the same context,assembly and resolution before
uniting data sets.
* validity function checking the format of the data for a methylRaw object is implemented
* clusterSamples() and PCASamples() take methylation context information automatically and use it in plot titles
methylKit 0.3.1
--------------
IMPROVEMENTS AND BUG FIXES
* syntax error when fisher.test applied at calculateMethDiff() is fixed.