[go: up one dir, main page]

Menu

[r49]: / trunk / command_line.txt  Maximize  Restore  History

Download this file

333 lines (219 with data), 10.0 kB

  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
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
{NORMAL}
{NORMAL_REQUIRED}
[--help]
Prints this help message.
# either one of these two
# why would config file need number of alleles?
# try whether we can remove it
[(-c|--configFile) <configFile>]
The input configuration file. Tells you meta parameter and which
sequences go into which demes.
[(-n|--numPerDeme) <numPerDeme>]
The number of individuals per deme. e.g. 5,3,2 means the first 5
haplotypes are in deme 0, the next 3 are in deme 1, and the next 2 are
in deme 2.
[--vcfFile <vcfFile>]
The sequence file. Just a bunch of sequences. In vcf format.
[--compositeLikelihood <compositeLikelihood>]
The composite likelihood type (LOL, PAC, SuperPAC, PCL, PCLOL, or FILE).
(-p|--paramFile) <paramFile>
The input parameter file.
# maybe break instead of ignoring
# ignoring is ok
[--vcfDiAllelic]
If this switch is set, the vcf file is represented with two alleles
internally. Sites with more then two alleles are ignored.
# maybe break instead of ignoring
# ignoring is ok
[--vcfTriAllelic]
If this switch is set, the vcf file is represented with three alleles
internally. Sites with more then three alleles are ignored.
# hmm, I guess we need this
[--vcfFilterPassString <vcfFilterPassString>]
The String in the FILTER column of a vcf-file that marks a snp as having
passed the filters. ALL OTHER SITES ARE IGNORED
# also good to have
[--vcfOffset <vcfOffset>]
Offset(s) to shift the positions given in the vcf file(s).
[--demoFile <demoFile>]
The demography parameter file.
[--bounds <bounds>]
A string of semi-colon separated list of pairs of doubles, each pair
spearated by a comma. This list gives the bounds for the parameters
during the EM-estimation.
# how to deal with when not needed?
[--ratesFile <ratesFile>]
The file with the exponential growth rates. Has to kind of match the
demogrpahy file.
# good
[--lociPerHmmStep <lociPerHmmStep>]
If this flag is used with k > 1, then each HMM-step consists of k loci,
and we assume there are no recombinations within blocks of k loci. Not
compatible with locus skipping. (default: 1)
# i would say make logUniform default, but it needs parameters
# see what the options are, and then maybe make a default
# make some of these required.
[--intervalType <intervalType>]
The method for creating the intervals.
[single,simple,meanrate,mixtureuniform,old]
[--intervalParams <intervalParams>]
The parameters for generating the intervals. [single=<none>,
simple=<none>, meanrate=<numIntervals>, mixtureuniform=<numIntervals>,
old=<numIntervals>] (default: )
[--printIntervals]
If this switch is set, the intervals used are printed.
# highlight EXCLUDED
[--bedFile <bedFile>]
Include *.bed file(s) that lists all regions of the chromosome that
should be excluded from analysis
[--metaStartFile <metaStartFile>]
A file containing the start points to be used in tab-separated format.
[--metaNumStartPoints <metaNumStartPoints>]
The number of starting points (in each dimension if we have a grid).
[--metaNumIterations <metaNumIterations>]
The number of meta optimization steps for the gentic algorithm.
(default: 1)
[--metaParallelEmSteps <metaParallelEmSteps>]
Number of EM steps to be exectued in parallel during gentic algorithm.
(default: 1)
[--metaKeepBest <metaKeepBest>]
Number of best points to keep after a meta step.
[--metaNumPoints <metaNumPoints>]
Number of points to be used in each meta iteration.
[--metaGridStart]
If flag is set we start with a grid of points, otherwise the start
points are chosen randomly.
[-v|--verbose]
Verbose for the EM procedure.
# make sure it's only required when actual randomness involved
--seed <seed>
The seed to initialize the randomness.
[--numPermutations <numPermutations>]
Number of permutations to use for PAC-like methods.
[--numCsdsPerPerm <numCsdsPerPerm>]
Number of CSDs used for each permutation.
[--diffPermsPerChunk]
If this switch is set, use different set of permutations for each
independent chunk.
[--permutationsFile <permutationsFile>]
Specfiy file(s) containing a list of permutations to be used to analyze
each chunk respectively.
[--startPoint <startPoint>]
The starting point for the EM.
[--numberIterationsEM <numberIterationsEM>]
EM terminates after the given number of iterations.
[--numberIterationsMstep <numberIterationsMstep>]
M-step search terminates after given number of iterations
[--parallel <parallel>]
Specifies number of parallel threads
[--printEmPath]
If this switch is set, we print the parameter values at each EM step.
# make this default
[--coordinateWiseMStep]
Do a coordinatewise M Step optimization. The number of iterations and
the reletive error thresholds are applied coordinatewise
# here is hidden randomness, not nice
[--coordinateOrder <coordinateOrder>]
Order to update parameters for coordinateWiseMStep. If not specified,
use random order.
{HIDDEN}
# leave it
[(-s|--sequenceFile) <sequenceFile>]
The sequence file. Just a bunch of sequences.
# leave it
[--unCompressedSequence]
If this switch is set, the sequence file is uncompressed and contains
full haplotypes.
# these have to do with non-VCF
[--alleles <alleles>]
A comma separated string of alleles, e.g. 'A,C,G,T'.
# these have to do with non-VCF
[--missingAlleles <missingAlleles>]
A comma separated string of the alleles that are interpreted as missing,
e.g. 'N,T,S'.
# dont even know whether this works
[--stationaryForPartiallyMissing]
If this switch is set, the partially missing sites are not ignored, but
the stationary distribution is used for them. Kinda only works with the
multilocus hmm version.
# make this standard and a hidden modifier to not do this
[--renormalizeMutationMatrix]
If this switch is set, we renormalize the mutation matrix so rows sum to
1
# not really sure what this one does
[--makeUniformMutationMatrix]
If this switch is set, we use a uniform mutation matrix instead of the
given one, and adjust the mutation rate accordingly.
# make a good default (0.005)
[--minimalMigration <minimalMigration>]
If this factor is given, you can always migrate anywhere with a minial
rate of min(theta,rho)*<factor>. (Except the rates you estimate)
# leave it
[--useLocusSkipping]
If this switch is set, we use the locus-skipping speedup (requires
uniform PIM mutations)
# leave it
[--ancientDemeStates]
If this switch is set, the diCal hidden states are a triplet (epoch,
ancientDeme, haplotype).
# make migratingEthan default
--trunkStyle <trunkStyle>
The style of trunk. [simple,oldCake,meanCake,multiCake]
# make a good default, probably 'average'
[--cakeStyle <cakeStyle>]
Where in the interval to get the cake factor
[beginning,middle,end,average] (default: null)
# what even
[--printLoci <printLoci>]
If this flag is specified with an integer k, additionally decodes at
every k loci, so that we can print at them.
[--printSnps]
Print at the SNPs as well.
# w00t?
[--addTrunkIntervals <addTrunkIntervals>]
Number of trunk intervals to add. (default: 0)
[--oldCakeRenormalizeSettings]
Use old cake intervals and don't renormalize
[--oldCompressedDataFormat]
In old data format, the compressed sequence file didn't know how many
total loci there were, and this info was gotten from the config file.
# make a good default, probably 0.2
[--nmFraction <nmFraction>]
The initial stepsize of the nelder mead optimizer is this fraction of
the starting point.
# just leave it
[--condOnTransitionType]
If this switch is set, we condition on the type of transition
(recombination or no recombination).
[--marginalKL]
If this switch is set, instead of using EM objective function, we use KL
divergence to average marginal posterior decoding.
# too complicated
[--metaStretchProportion <metaStretchProportion>]
Probability that a new point is chosen in a ridge-like manner, as
opposed to each dimension independently. (default: 0.25)
[--metaDisperseFactor <metaDisperseFactor>]
SD for new Gaussian points is multiplied by this factor. (default: 1)
[--estimateRecomScaling]
If this switch is set, the EM will estimate a scaling factor for the
recombination rate, so inputRecomRate * scaling = recomRate.
[--estimateTheta]
If this switch is set, the EM will estimate theta for each interval
[--thetaIntervals <thetaIntervals>]
factory type by which to build the theta intervals, recommend
customFixed
[--thetaIntervalParams <thetaIntervalParams>]
params for the thetaIntervals factory.
# these didn't work anyways
[--relativeErrorM <relativeErrorM>]
M-step optimization terminates if relative improvement is less then the
given value.
[--relativeErrorE <relativeErrorE>]
EM optimization terminates if relative likelihood improvement is less
then the given value.
[--useParamRelErr]
If this switch is set, we use the relative error of the parameters
(instead of log-likelihood) as stopping criteria of E-step.
[--useParamRelErrM]
If this switch is set, we use the relative error of parameters (instead
of log-likelihood) as stopping criteria for M-step.