-
-
Notifications
You must be signed in to change notification settings - Fork 26
/
internal.r
2986 lines (2840 loc) · 106 KB
/
internal.r
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
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
#
# This software was authored by Zhian N. Kamvar and Javier F. Tabima, graduate
# students at Oregon State University; Jonah C. Brooks, undergraduate student at
# Oregon State University; and Dr. Nik Grünwald, an employee of USDA-ARS.
#
# Permission to use, copy, modify, and distribute this software and its
# documentation for educational, research and non-profit purposes, without fee,
# and without a written agreement is hereby granted, provided that the statement
# above is incorporated into the material, giving appropriate attribution to the
# authors.
#
# Permission to incorporate this software into commercial products may be
# obtained by contacting USDA ARS and OREGON STATE UNIVERSITY Office for
# Commercialization and Corporate Development.
#
# The software program and documentation are supplied "as is", without any
# accompanying services from the USDA or the University. USDA ARS or the
# University do not warrant that the operation of the program will be
# uninterrupted or error-free. The end-user understands that the program was
# developed for research purposes and is advised not to rely exclusively on the
# program for any reason.
#
# IN NO EVENT SHALL USDA ARS OR OREGON STATE UNIVERSITY BE LIABLE TO ANY PARTY
# FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING
# LOST PROFITS, ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION,
# EVEN IF THE OREGON STATE UNIVERSITY HAS BEEN ADVISED OF THE POSSIBILITY OF
# SUCH DAMAGE. USDA ARS OR OREGON STATE UNIVERSITY SPECIFICALLY DISCLAIMS ANY
# WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
# MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE AND ANY STATUTORY
# WARRANTY OF NON-INFRINGEMENT. THE SOFTWARE PROVIDED HEREUNDER IS ON AN "AS IS"
# BASIS, AND USDA ARS AND OREGON STATE UNIVERSITY HAVE NO OBLIGATIONS TO PROVIDE
# MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
#
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
#==============================================================================#
# This function will attempt to convert external files of the following types:
#
# Structure (*.str, *.stru)
# Fstat (*.dat)
# Gentix (*.gtx)
# Genpop (*.gen)
# Genalex (*.csv)
#
# The output is a poppr object and the original filename that was imported.
# Missing data is handled via the function missingno and the ability for clone
# correction is also possible.
# If quiet is set to false and you are importing non-genind or poppr objects,
# you will see many warnings.
#
# Public functions utilizing this function:
# # poppr
#
# Internal functions utilizing this function:
# # new.poppr (in testing)
#==============================================================================#
process_file <- function(input, quiet=TRUE, missing="ignore", cutoff=0.05,
keep=1, clonecorrect=FALSE, strata=1){
if (!is.genind(input)){
x <- input
if (toupper(.readExt(x)) == "CSV"){
try(input <- read.genalex(x), silent=quiet)
try(input <- read.genalex(x, region=TRUE), silent=quiet)
try(input <- read.genalex(x, geo=TRUE), silent=quiet)
try(input <- read.genalex(x, geo=TRUE, region=TRUE), silent=quiet)
} else {
try(input <- import2genind(x, quiet=quiet), silent=quiet)
}
stopifnot(is.genind(input))
input@call[2] <- x
popcall <- input@call
input <- missingno(input, type=missing, cutoff=cutoff, quiet=quiet)
input@call <- popcall
if (clonecorrect == TRUE){
poplist <- clonecorrect(input, strata = strata, keep = keep)
input <- poplist
input@call <- popcall
}
} else if (is.genind(input)) {
x <- as.character(match.call()[2])
popcall <- input@call
input <- missingno(input, type=missing, cutoff=cutoff, quiet=quiet)
if (clonecorrect == TRUE){
poplist <- clonecorrect(input, strata = strata, keep = keep)
input <- poplist
input@call <- popcall
}
}
return(list(X=x, GENIND=input))
}
#==============================================================================#
# .clonecorrector will simply give a list of individuals (rows) that are
# duplicated within a genind object. This can be used for clone correcting a
# single genind object.
#
# Public functions utilizing this function:
# # clonecorrect, bruvo.msn
#
# Internal functions utilizing this function:
# # none
#==============================================================================#
.clonecorrector <- function(x){
if (is.genclone(x) | is(x, "snpclone")){
is_duplicated <- duplicated(x@mlg[])
} else {
is_duplicated <- duplicated(x@tab[, 1:ncol(x@tab)])
}
res <- -which(is_duplicated)
# conditional for the case that all individuals are unique.
if(is.na(res[1])){
res <- which(!is_duplicated)
}
return(res)
}
#==============================================================================#
# This will remove either loci or genotypes containing missing values above the
# cutoff percent.
#
# Public functions utilizing this function:
# # missingno
#
# Internal functions utilizing this function:
# # none.
#==============================================================================#
percent_missing <- function(pop, type="loci", cutoff=0.05){
if (toupper(type) == "LOCI"){
missing_loci <- 1 - propTyped(pop, "loc")
locfac <- locFac(pop)
names(missing_loci) <- levels(locfac)
missing_loci <- missing_loci[missing_loci <= cutoff]
misslist <- 1:length(locfac)
filter <- locfac %in% names(missing_loci)
} else {
missing_geno <- 1 - propTyped(pop, "ind")
misslist <- 1:nInd(pop)
filter <- missing_geno <= cutoff
}
return(misslist[filter])
}
#==============================================================================#
# This implements rounding against the IEEE standard and rounds 0.5 up
# Public functions utilizing this function:
# # none
#
# Internal functions utilizing this function:
# # .PA.pairwise.differences, .pairwise.differences
#==============================================================================#
round.poppr <- Vectorize(function(x){
ix <- as.integer(x)
is_even <- ix %% 2 == 0
if (is_even) {
if (x - ix == 0.5)
x <- round(x) + 1
else if (-x + ix == 0.5)
x <- round(x) - 1
} else {
x <- round(x)
}
return(x)
})
#==============================================================================#
# Subsetting the population and returning the indices.
#
# Public functions utilizing this function:
# ## mlg.crosspop poppr.msn
#
# Internal functions utilizing this function:
# ## none
#==============================================================================#
sub_index <- function(pop, sublist="ALL", blacklist=NULL){
numList <- seq(nInd(pop))
if (is.null(pop(pop))){
return(numList)
}
if(toupper(sublist[1]) == "ALL"){
if (is.null(blacklist)){
return(numList)
} else {
# filling the sublist with all of the population names.
sublist <- popNames(pop)
}
}
# Treating anything present in blacklist.
if (!is.null(blacklist)){
# If both the sublist and blacklist are numeric or character.
if (is.numeric(sublist) & is.numeric(blacklist) | class(sublist) == class(blacklist)){
sublist <- sublist[!sublist %in% blacklist]
} else if (is.numeric(sublist) & class(blacklist) == "character"){
# if the sublist is numeric and blacklist is a character. eg s=1:10, b="USA"
sublist <- sublist[sublist %in% which(!popNames(pop) %in% blacklist)]
} else {
# no sublist specified. Ideal situation
if(all(popNames(pop) %in% sublist)){
sublist <- sublist[-blacklist]
} else {
# weird situation where the user will specify a certain sublist, yet index
# the blacklist numerically. Interpreted as an index of populations in the
# whole data set as opposed to the sublist.
warning("Blacklist is numeric. Interpreting blacklist as the index of the population in the total data set.")
sublist <- sublist[!sublist %in% popNames(pop)[blacklist]]
}
}
}
# subsetting the population.
if (is.numeric(sublist)){
sublist <- popNames(pop)[sublist]
} else{
sublist <- popNames(pop)[popNames(pop) %in% sublist]
}
# sublist <- (1:length(pop@pop))[pop@pop %in% sublist]
sublist <- pop(pop) %in% sublist
if (sum(sublist) == 0){
warning("All items present in Sublist are also present in the Blacklist.\nSubsetting not taking place.")
return(seq(nInd(pop)))
}
#cat("Sublist:\n", sublist,"\n")
return(numList[sublist])
}
#==============================================================================#
# Internal function to create mlg.table.
#
# Public functions utilizing this function:
# # mlg.table mlg.crosspop
#
# Internal functions utilizing this function:
# # none
#==============================================================================#
mlg.matrix <- function(x){
visible <- "original"
if (is.genclone(x) | is(x, "snpclone")){
mlgvec <- x@mlg[]
if (is(x@mlg, "MLG")){
visible <- visible(x@mlg)
}
} else {
mlgvec <- mlg.vector(x)
}
if (!is.null(pop(x))){
mlg.mat <- table(pop(x), mlgvec)
} else {
mlg.mat <- t(as.matrix(table(mlgvec)))
rownames(mlg.mat) <- "Total"
}
names(attr(mlg.mat, "dimnames")) <- NULL
if (visible == "custom"){
return(mlg.mat)
}
if (is.null(colnames(mlg.mat))){
mlgs <- length(unique(mlgvec))
colnames(mlg.mat) <- 1:mlgs
}
colnames(mlg.mat) <- paste("MLG", colnames(mlg.mat), sep=".")
return(unclass(mlg.mat))
}
# #==============================================================================#
# # DEPRECATED
# #==============================================================================#
# old.mlg.matrix <- function(x){
# mlgvec <- mlg.vector(x)
# mlgs <- length(unique(mlgvec))
#
# if (!is.null(x@pop)){
# # creating a new population matrix. Rows are the population indicator and
# # columns are the genotype indicator.
# mlg.mat <- matrix(ncol=mlgs, nrow=length(levels(x@pop)), data=0L)
# # populating (no, pun intended.) the matrix with genotype counts.
# lapply(levels(x@pop), function(z){
# # This first part gets the index for the row names.
# count <- as.numeric(substr(z, 2, nchar(z)))
# sapply(mlgvec[which(x@pop==z)],
# function(a) mlg.mat[count, a] <<-
# mlg.mat[count, a] + 1L)
# })
# rownames(mlg.mat) <-popNames(x)
# } else {
# # if there are no populations to speak of.
# mlg.mat <- t(as.matrix(vector(length=mlgs, mode="numeric")))
# sapply(mlgvec, function(a) mlg.mat[a] <<- mlg.mat[a] + 1)
# rownames(mlg.mat) <- "Total"
# }
# colnames(mlg.mat) <- paste("MLG", 1:mlgs, sep=".")
# return(mlg.mat)
# }
#==============================================================================#
# !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! #
#
# The reason for this section of code is for the fact that Presence/Absence
# markers are dealt with in a different way for adegenet (to save memory) and
# so the calculations must be different as implemented in these mostly identical
# functions.
#
# Public functions utilizing this function:
# # ia
#
# Internal functions utilizing this function:
# # .ia
#
# !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! #
#==============================================================================#
#' @noRd
.PA.Ia.Rd <- function(pop, missing=NULL){
vard.vector <- NULL
numLoci <- ncol(pop@tab)
numIsolates <- nrow(pop@tab)
# Creating this number is necessary because it is how the variance is
# calculated.
np <- choose(numIsolates, 2)
if(np < 2){
return(as.numeric(c(NaN, NaN)))
}
# Starting the actual calculations.
V <- .PA.pairwise.differences(pop, numLoci, np, missing=missing)
# First, set the variance of D
varD <- ((sum(V$D.vector^2)-((sum(V$D.vector))^2)/np))/np
# Next is to create a vector containing all of the variances of d (there
# will be one for each locus)
vard.vector <- ((V$d2.vector-((V$d.vector^2)/np))/np)
vardpair.vector <- .Call("pairwise_covar", vard.vector, PACKAGE = "poppr")
# The sum of the variances necessary for the calculation of Ia is calculated
sigVarj <- sum(vard.vector)
rm(vard.vector)
# Finally, the Index of Association and the standardized Index of associati-
# on are calculated.
Ia <- (varD/sigVarj)-1
rbarD <- (varD - sigVarj)/(2*sum(vardpair.vector))
return(c(Ia, rbarD))
}
#==============================================================================#
# .PA.pairwise.differences will calculate three vectors that will be used for the
# calculation of the Index of Association and standardized Index of Association
# Later.
# pop = genind object
# numLoci = should read numLoci. This will be fixed later.
# temp.d.vector = temporary vector to store the differences
# d.vector = a vector of the sum of the differences at each locus. The length
# of this vector will be the same as the number of loci.
# d2.vector = the same as d.vector, except it's the sum of the squares
# D.vector = a vector of the the pairwise distances over all loci. The length
# of this vector will be the same as n(n-1)/2, where n is number of
# isolates.
# Public functions utilizing this function:
# # ia
#
# Internal functions utilizing this function:
# # .ia
#
#==============================================================================#
.PA.pairwise.differences <- function(pop, numLoci, np, missing){
temp.d.vector <- matrix(nrow = np, ncol = numLoci, data = NA_real_)
if( missing == "MEAN" ){
# this will round all of the values if the missing indicator is "mean"
temp.d.vector <- vapply(seq(numLoci),
function(x) as.vector(dist(pop@tab[, x])),
temp.d.vector[, 1])
# since the replacement was with "mean", the missing data will not produce
# a binary distance. The way we will handle this is to replace numbers that
# are not one or zero with a rounded value.
tempz <- !temp.d.vector %in% 0:1
temp.d.vector[tempz] <- vapply(temp.d.vector[tempz], round.poppr, 1)
} else {
temp.d.vector <- vapply(seq(numLoci),
function(x) as.vector(dist(pop@tab[, x])),
temp.d.vector[, 1])
# checking for missing data and imputing the comparison to zero.
if(any(is.na(temp.d.vector))){
temp.d.vector[which(is.na(temp.d.vector))] <- 0
}
}
if (max(ploidy(pop)) > 1){
# multiplying by two is the proper way to evaluate P/A diploid data because
# one cannot detect heterozygous loci (eg, a difference of 1).
temp.d.vector <- temp.d.vector * max(ploidy(pop))
d.vector <- as.vector(colSums(temp.d.vector))
d2.vector <- as.vector(colSums(temp.d.vector^2))
D.vector <- as.vector(rowSums(temp.d.vector))
} else {
d.vector <- as.vector(colSums(temp.d.vector))
d2.vector <- d.vector
D.vector <- as.vector(rowSums(temp.d.vector))
}
vectors <- list(d.vector = d.vector, d2.vector = d2.vector, D.vector = D.vector)
return(vectors)
}
#==============================================================================#
# Function for parsing output of poppr function.
#
# Public functions utilizing this function:
# # poppr, poppr.all
#
# Internal functions utilizing this function:
# # none
#
#==============================================================================#
final <- function(Iout, result){
if (is.null(result)){
return(Iout)
} else {
return(result)
}
}
#==============================================================================#
# The internal version of ia.
# Public functions utilizing this function:
# # ia, poppr
#
# Internal functions utilizing this function:
# # none
#
#==============================================================================#
.ia <- function(pop, sample=0, method=1, quiet=FALSE, namelist=NULL,
missing="ignore", hist=TRUE, index = "rbarD"){
METHODS = c("permute alleles", "parametric bootstrap",
"non-parametric bootstrap", "multilocus")
if(pop@type!="PA"){
type <- pop@type
popx <- seploc(pop)
}
else {
type <- pop@type
popx <- pop
.Ia.Rd <- .PA.Ia.Rd
}
# if there are less than three individuals in the population, the calculation
# does not proceed.
if (nInd(pop) < 3){
IarD <- as.numeric(c(NA, NA))
names(IarD) <- c("Ia", "rbarD")
if(sample==0){
return(IarD)
}
else{
IarD <- as.numeric(rep(NA, 4))
names(IarD) <- c("Ia","p.Ia","rbarD","p.rD")
return(IarD)
}
}
IarD <- .Ia.Rd(popx, missing)
# data vomit options.
if (!quiet){
cat("|", namelist$population ,"\n")
}
names(IarD) <- c("Ia", "rbarD")
# no sampling, it will simply return two named numbers.
if (sample == 0){
Iout <- IarD
result <- NULL
}
# sampling will perform the iterations and then return a data frame indicating
# the population, index, observed value, and p-value. It will also produce a
# histogram.
else{
Iout <- NULL
idx <- as.data.frame(list(Index=names(IarD)))
samp <- .sampling(popx, sample, missing, quiet=quiet, type=type, method=method)
p.val <- sum(IarD[1] <= c(samp$Ia, IarD[1]))/(sample + 1)#ia.pval(index="Ia", samp2, IarD[1])
p.val[2] <- sum(IarD[2] <= c(samp$rbarD, IarD[2]))/(sample + 1)#ia.pval(index="rbarD", samp2, IarD[2])
if(hist == TRUE){
print(poppr.plot(samp, observed=IarD, pop=namelist$population, index = index,
file=namelist$File, pval=p.val, N=nrow(pop@tab)))
}
result <- 1:4
result[c(1, 3)] <- IarD
result[c(2, 4)] <- p.val
names(result) <- c("Ia","p.Ia","rbarD","p.rD")
iaobj <- list(index = final(Iout, result), samples = samp)
class(iaobj) <- "ialist"
return(iaobj)
}
return(final(Iout, result))
}
#==============================================================================#
#==============================================================================#
#=====================Index of Association Calculations========================#
#==============================================================================#
#==============================================================================#
# .pairwise.differences will calculate three vectors that will be used for the
# calculation of the Index of Association and standardized Index of Association
# Later. Note that all NA's must be treated or removed before this step.
# pop = genind object
# numLoci = should read numLoci. This will be fixed later.
# temp.d.vector = temporary vector to store the differences
# d.vector = a vector of the sum of the differences at each locus. The length
# of this vector will be the same as the number of loci.
# d2.vector = the same as d.vector, except it's the sum of the squares
# D.vector = a vector of the the pairwise distances over all loci. The length
# of this vector will be the same as n(n-1)/2, where n is number of
# isolates.
#
#
# # DEPRECATED
# #==============================================================================#
# .pairwise.differences <- function(pop, numLoci, np, missing){
# temp.d.vector <- matrix(nrow=np, ncol=numLoci, data=as.numeric(NA))
# if( missing == "MEAN" )
# temp.d.vector <- matrix(nrow=np, ncol=numLoci,
# data=vapply(vapply(pop, pairwisematrix,
# temp.d.vector[, 1], np), round.poppr, 1))
# else
# temp.d.vector <- vapply(pop, pairwisematrix, temp.d.vector[, 1], np)
# d.vector <- as.vector(colSums(temp.d.vector))
# d2.vector <- as.vector(colSums(temp.d.vector^2))
# D.vector <- as.vector(rowSums(temp.d.vector))
# vectors <- list(d.vector=d.vector, d2.vector=d2.vector, D.vector=D.vector)
# return(vectors)
# }
#==============================================================================#
# pairwisematrix performs a pairwise comparison over all individuals per locus
# and returns a vector that will make its way into the final matrix for d.
# the conditional for this is that each locus must not be completely fixed for
# one allele. In that case, the resulting pairwise differences will all be zero.
#
#
# DEPRECATED
#==============================================================================#
# pairwisematrix <- function(pop, np){
# temp.d.vector <- vector(mode="numeric", length=np)
# if ( ncol(pop@tab) != 1 )
# temp.d.vector <- as.numeric(colSums(.pairwise.diffs(t(pop@tab)), na.rm=TRUE))
# return(temp.d.vector)
# }
#==============================================================================#
# The original function pairwise.diffs can be found here
# https://stat.ethz.ch/pipermail/r-help/2004-August/055324.html
#
#
# DEPRECATED
#==============================================================================#
# .pairwise.diffs <- function(x){
# stopifnot(is.matrix(x))
#
# # create column combination pairs
# prs <- cbind(rep(1:ncol(x), each = ncol(x)), 1:ncol(x))
# col.diffs <- prs[prs[, 1] < prs[, 2], , drop = FALSE]
#
# # do pairwise differences
# result <- abs(x[, col.diffs[, 1]] - x[, col.diffs[, 2], drop = FALSE])
#
# return(result)
# }
#==============================================================================#
# To calculate rbarD, the pairwise variances for each locus needs to be
# calculated.
#
#
# DEPRECATED
#==============================================================================#
# .pairwise.variances <- function(vard.vector, pair.alleles){
# # Here the roots of the products of the variances are being produced and
# # the sum of those values is taken.
# vardpair.vector <- vector(length=pair.alleles)
# vardpair.vector <- sqrt(combn(vard.vector, 2, prod))
# return(vardpair.vector)
# }
#==============================================================================#
# The actual calculation of Ia and rbarD. This allows for multiple populations
# to be calculated.
# pop: A list of genind objects consisting of one locus each over a population.
# Public functions utilizing this function:
# # none
#
# Internal functions utilizing this function:
# # .ia
#
#==============================================================================#
.Ia.Rd <- function (pop, missing = NULL)
{
vard.vector <- NULL
numLoci <- length(pop)
numIsolates <- nInd(pop[[1]])
np <- choose(numIsolates, 2)
if (np < 2) {
return(as.numeric(c(NaN, NaN)))
}
V <- pair_diffs(pop, numLoci, np)
varD <- ((sum(V$D.vector^2) - ((sum(V$D.vector))^2)/np))/np
vard.vector <- ((V$d2.vector - ((V$d.vector^2)/np))/np)
vardpair.vector <- .Call("pairwise_covar", vard.vector, PACKAGE = "poppr")
sigVarj <- sum(vard.vector)
rm(vard.vector)
Ia <- (varD/sigVarj) - 1
rbarD <- (varD - sigVarj)/(2 * sum(vardpair.vector))
return(c(Ia, rbarD))
}
#==============================================================================#
# This creates the results from the pairwise difference matrix used by .Ia.Rd
# to calculate the index of association.
#
# Public functions utilizing this function:
# # none
#
# Internal functions utilizing this function:
# # .Ia.Rd
#
#==============================================================================#
pair_diffs <- function(pop, numLoci, np)
{
temp.d.vector <- pair_matrix(pop, numLoci, np)
d.vector <- colSums(temp.d.vector)
d2.vector <- colSums(temp.d.vector^2)
D.vector <- rowSums(temp.d.vector)
return(list(d.vector = d.vector, d2.vector = d2.vector, D.vector = D.vector))
}
#==============================================================================#
# This creates a pairwise difference matrix via the C function pairdiffs in
# src/poppr_distance.c
#
# Public functions utilizing this function:
# # none
#
# Internal functions utilizing this function:
# # pair_diffs, pair.ia
#
#==============================================================================#
pair_matrix <- function(pop, numLoci, np)
{
temp.d.vector <- matrix(nrow = np, ncol = numLoci, data = as.numeric(NA))
temp.d.vector <- vapply(pop, function(x) .Call("pairdiffs", tab(x),
PACKAGE = "poppr")/2,
FUN.VALUE = temp.d.vector[, 1])
temp.d.vector <- ceiling(temp.d.vector)
return(temp.d.vector)
}
#==============================================================================#
# Internal counter...probably DEPRECATED.
#==============================================================================#
# .new_counter <- function() {
# i <- 0
# function() {
# i <<- i + 1
# i
# }
# }
#==============================================================================#
# Bruvo's distance calculation that takes in an SSR matrix. Note the conditions
# below.
#
# Public functions utilizing this function:
# # bruvo.boot
#
# Internal functions utilizing this function:
# # none
#
# DEPRECATED
#==============================================================================#
# phylo.bruvo.dist <- function(ssr.matrix, replen=c(2), ploid=2, add = TRUE, loss = TRUE){
# # Preceeding functions should take care of this:
# # ssr.matrix <- genind2df(pop, sep="/", usepop=FALSE)
# # ssr.matrix[is.na(ssr.matrix)] <- paste(rep(0, ploid), collapse="/")
# # Bruvo's distance needs a matrix with the number of columns equal to the
# # number of loci multiplied by the polidy.
# indnames <- rownames(ssr.matrix)
# ssr.matrix <- apply(ssr.matrix, 1, strsplit, "/")
# # Getting the values into numeric form.
# ssr.matrix <- apply(as.matrix(t(sapply(ssr.matrix, unlist))), 2, as.numeric)
# # Dividing each column by the repeat length and changing the values to integers.
# ssr.matrix <- apply(ssr.matrix / rep(replen, each=ploid*nrow(ssr.matrix)), 2, round)
# ssr.matrix <- apply(ssr.matrix, 2, as.integer)
# perms <- .Call("permuto", ploid, PACKAGE = "poppr")
# distmat <- .Call("bruvo_distance", ssr.matrix, perms, ploid, add, loss, PACKAGE = "poppr")
# distmat[distmat == 100] <- NA
# avg.dist.vec <- apply(distmat, 1, mean, na.rm=TRUE)
# # presenting the information in a lower triangle distance matrix.
# dist.mat <- matrix(ncol=nrow(ssr.matrix), nrow=nrow(ssr.matrix))
# dist.mat[which(lower.tri(dist.mat)==TRUE)] <- avg.dist.vec
# dist.mat <- as.dist(dist.mat)
# attr(dist.mat, "labels") <- indnames
# return(dist.mat)
# }
#==============================================================================#
# This will transform the data to be in the range of [0, 1]
#
# Public functions utilizing this function:
# poppr.msn
#
# Internal functions utilizing this function:
# # adjustcurve
#==============================================================================#
rerange <- function(x){
minx <- min(x, na.rm = TRUE)
maxx <- max(x, na.rm = TRUE)
if (!is.finite(minx) || !is.finite(maxx)){
warning("non-finite values found for distances, returning 0.5")
return(rep(0.5, length(x)))
}
if (minx < 0)
x <- x + abs(minx)
maxx <- maxx + abs(minx)
if (maxx > 1)
x <- x/maxx
return(x)
}
#==============================================================================#
# This will scale the edge widths
#
# Public functions utilizing this function:
# none
#
# Internal functions utilizing this function:
# # update_edge_scales
#==============================================================================#
make_edge_width <- function(mst){
edgewidth <- rerange(E(mst)$weight)
if (any(edgewidth < 0.08)){
edgewidth <- edgewidth + 0.08
}
return(1/edgewidth)
}
#==============================================================================#
# This will scale the edge widths and edge color for a graph
#
# Public functions utilizing this function:
# poppr.msn bruvo.msn plot_poppr_msn
#
# Internal functions utilizing this function:
# # singlepop_msn
#==============================================================================#
update_edge_scales <- function(mst, wscale = TRUE, gscale = TRUE, glim, gadj){
if(gscale == TRUE){
E(mst)$color <- gray(adjustcurve(E(mst)$weight, glim=glim, correction=gadj,
show=FALSE))
} else {
E(mst)$color <- rep("black", length(E(mst)$weight))
}
E(mst)$width <- 2
if (wscale==TRUE){
E(mst)$width <- make_edge_width(mst)
}
return(mst)
}
#==============================================================================#
# This will adjust the grey scale with respect to the edge weights for igraph.
# This is needed because the length of the edges do not correspond to weights.
# If show is set to TRUE, it will show a graph giving the equation used for con-
# version from the original scale to the grey scale, the grey scale itself in
# the background, and the curve.
#
# Public functions utilizing this function:
# # bruvo.msn (soon)
#
# Internal functions utilizing this function:
# # new.bruvo.msn
# # new.poppr.msn
#
#==============================================================================#
adjustcurve <- function(weights, glim = c(0, 0.8), correction = 3, show=FALSE,
scalebar = FALSE, smooth = TRUE){
w <- weights
w <- rerange(w)
maxg <- max(glim)
ming <- 1-(min(glim)/maxg)
if (correction < 0){
adj <- (w^abs(correction))/(1/ming)
adj <- (adj + 1-ming) / ((1 / maxg))
} else {
adj <- (1 - (((1-w)^abs(correction))/(1/ming)) )
adj <- adj / (1/maxg)
}
if (!show){
return(adj)
} else if (!scalebar){
with_quantiles <- sort(weights)
wq_raster <- t(as.raster(as.matrix(gray(sort(adj)), nrow = 1)))
xlims <- c(min(weights), max(weights))
graphics::plot(xlims, 0:1, type = "n", ylim = 0:1, xlim = xlims, xlab = "", ylab = "")
graphics::rasterImage(wq_raster, xlims[1], 0, xlims[2], 1)
graphics::points(x = sort(weights), y = sort(adj), col = grDevices::grey(rev(sort(adj))), pch=20)
title(xlab="Observed Value", ylab="Grey Adjusted",
main=paste("Grey adjustment\n min:",
min(glim),
"max:", max(glim),
"adjust:", abs(correction)))
if (correction < 0){
graphics::text(bquote(frac(bgroup("(", frac(scriptstyle(x)^.(abs(correction)),
.(ming)^-1),")") + .(1-ming),
.(maxg)^-1)) ,
x = min(weights) + (0.25*max(weights)), y=0.75, col="red")
} else {
graphics::text(bquote(frac(1-bgroup("(", frac((1-scriptstyle(x))^.(abs(correction)),
.(ming)^-1),")"),
.(maxg)^-1)) ,
x= min(weights) + (0.15*max(weights)), y=0.75, col="red")
}
graphics::lines(x=xlims, y=c(min(glim), min(glim)), col="yellow")
graphics::lines(x=xlims, y=c(max(glim), max(glim)), col="yellow")
} else {
with_quantiles <- sort(weights)
wq_raster <- t(grDevices::as.raster(as.matrix(grDevices::gray(sort(adj)), nrow = 1)))
no_quantiles <- seq(min(weights), max(weights), length = 1000)
nq_raster <- adjustcurve(no_quantiles, glim, correction, show = FALSE)
nq_raster <- t(grDevices::as.raster(as.matrix(grDevices::gray(nq_raster), nrow = 1)))
graphics::layout(matrix(1:2, nrow = 2))
graphics::plot.new()
graphics::rasterImage(wq_raster, 0, 0.5, 1, 1)
graphics::polygon(c(0, 1, 1), c(0.5, 0.5, 0.8), col = "white", border = "white", lwd = 2)
graphics::axis(3, at = c(0, 0.25, 0.5, 0.75, 1), labels = round(quantile(with_quantiles), 3))
graphics::text(0.5, 0, labels = "Quantiles From Data", font = 2, cex = 1.5, adj = c(0.5, 0))
graphics::plot.new()
graphics::rasterImage(nq_raster, 0, 0.5, 1, 1)
graphics::polygon(c(0, 1, 1), c(0.5, 0.5, 0.8), col = "white", border = "white", lwd = 2)
graphics::axis(3, at = c(0, 0.25, 0.5, 0.75, 1), labels = round(quantile(no_quantiles), 3))
graphics::text(0.5, 0, labels = "Quantiles From Smoothing", font = 2, cex = 1.5, adj = c(0.5, 0))
# Return top level plot to defau lts.
graphics::layout(matrix(c(1), ncol=1, byrow=T))
graphics::par(mar=c(5, 4, 4, 2) + 0.1) # number of lines of margin specified.
graphics::par(oma=c(0, 0, 0, 0)) # Figure margins
}
}
#==============================================================================#
# This will guess the repeat lengths of the microsatellites for Bruvo's distance
#
# Public functions utilizing this function:
# # bruvo.boot bruvo.dist
#
# Internal functions utilizing this function:
# # none
#
#==============================================================================#
guesslengths <- function(vec){
if (length(vec) > 1){
lens <- vapply(2:length(vec), function(x) abs(vec[x] - vec[x - 1]), 1)
if (all(lens == 1)){
return(1)
} else {
return(min(lens[lens > 1]))
}
} else {
return(1)
}
}
#==============================================================================#
# Specifically used to find loci with an excessive number of the same genotype.
#
# Public functions utilizing this function:
# # informloci
#
# Internal functions utilizing this function:
# # none
#==============================================================================#
test_table <- function(loc, min_ind, n){
tab <- table(loc)
return(ifelse(any(tab > n - min_ind), FALSE, TRUE))
}
#==============================================================================#
# Normalize negative branch lenght by converting the negative branch to zero
# and adding the negative value to the sibling branch.
#
# This now has a few caveats:
# 1. If the parent branch contains a polytomy, then only the branch that is
# equal to the negative branch will be fixed.
# 2. If there are branches that cannot be fixed (i.e., the absolute value of
# the negative branch is greater than the value of the sibling), then it will
# not be fixed.
#
# Public functions utilizing this function:
# # bruvo.boot
#
# Internal functions utilizing this function:
# # none
#==============================================================================#
fix_negative_branch <- function(tre){
# Creating a matrix from the tree information: Tree edges and edge length
all.lengths <- matrix(c(tre$edge, tre$edge.length), ncol = 3,
dimnames = list(1:length(tre$edge.length),
c("parent", "child", "length")
))
# Looking at the edges that are zero.
zero.edges <- all.lengths[tre$edge.length < 0, , drop = FALSE]
# Checking which negative edges are included in all the edges
all.edges <- all.lengths[all.lengths[, "parent"] %in% zero.edges[, "parent"], , drop = FALSE]
# Ordering all the edges
index.table <- all.edges[order(all.edges[, "parent"]), , drop = FALSE]
# Loop to change the NJ branch length
for (i in (unique(index.table[, "parent"]))){
the_parents <- index.table[, "parent"] == i
fork <- index.table[, "length"][the_parents]
# Check for polytomies
if (length(fork) > 2){
# Fix only siblings of equal magnitude.
forkinds <- abs(fork) == -min(fork)
fork <- fork[forkinds]
fixed_lengths <- abs(fork) + min(fork)
# Check that branches are actually fixed.
if (all(fixed_lengths >= 0)){
index.table[, "length"][the_parents][forkinds] <- abs(fork) + min(fork)
}
} else { # No polytomies
fixed_lengths <- abs(fork) + min(fork)
# Check that branches are actually fixed.
if (all(fixed_lengths >= 0)){
index.table[, "length"][the_parents] <- abs(fork) + min(fork)
}
}
}
# replacing the branch length for each negative value in the total table
name_match <- match(rownames(index.table), rownames(all.lengths))
all.lengths[, "length"][name_match] <- index.table[, "length"]
# replacing the branch lengths to the original tree
tre$edge.length <- all.lengths[, "length"]
return(tre)
}
#==============================================================================#
# Calculate Bruvo's distance from a bruvomat object.
#
# Public functions utilizing this function:
# # bruvo.msn, bruvo.dist, bruvo.boot
#
# Internal functions utilizing this function:
# # singlepop_msn
#==============================================================================#
bruvos_distance <- function(bruvomat, funk_call = match.call(), add = TRUE,
loss = TRUE, by_locus = FALSE){
x <- bruvomat@mat
ploid <- bruvomat@ploidy
if (getOption("old.bruvo.model") && ploid > 2 && (add | loss)){
msg <- paste("The option old.bruvo.model has been set to TRUE, which does",
"not represent every ordered combinations of alleles in the",
"genome addition or loss models. This could result in",
"potentially incorrect results.",
"\n\n To use every ordered combination of alleles for",
"estimating short genotypes, enter the following command in",
"your R console:",
"\n\n\toptions(old.bruvo.model = FALSE)\n")
warning(msg, call. = FALSE, immediate. = TRUE)
}
replen <- bruvomat@replen
x[is.na(x)] <- 0
# Dividing the data by the repeat length of each locus.
x <- x / rep(replen, each = ploid * nrow(x))
x <- matrix(as.integer(round(x)), ncol = ncol(x))
# Getting the permutation vector.
perms <- .Call("permuto", ploid, PACKAGE = "poppr")
# Calculating bruvo's distance over each locus.
distmat <- .Call("bruvo_distance",
x, # data matrix
perms, # permutation vector (0-indexed)
ploid, # maximum ploidy
add, # Genome addition model switch
loss, # Genome loss model switch
getOption("old.bruvo.model"), # switch to use unordered genotypes
PACKAGE = "poppr")
# If there are missing values, the distance returns 100, which means that the
# comparison is not made. These are changed to NA.
distmat[distmat == 100] <- NA
if (!by_locus){
# Obtaining the average distance over all loci.
avg.dist.vec <- apply(distmat, 1, mean, na.rm=TRUE)
# presenting the information in a lower triangle distance matrix.
dist.mat <- matrix(ncol=nrow(x), nrow=nrow(x))
dist.mat[which(lower.tri(dist.mat)==TRUE)] <- avg.dist.vec
dist.mat <- as.dist(dist.mat)