-
Notifications
You must be signed in to change notification settings - Fork 1
/
Musa_endo_analyses.r
1784 lines (1508 loc) · 76.7 KB
/
Musa_endo_analyses.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
#####################################################
#####################################################
#####################################################
###### ######
###### Fungal Endophytes in Musa CWR Seeds ######
###### ######
#####################################################
#####################################################
#####################################################
library(ape)
library(colorspace)
library(dplyr)
library(egg)
library(eulerr)
library(ggplot2)
library(ggplotify)
library(ggpubr)
library(ggrepel)
library(ggstance)
library(ggtree)
library(ggwordcloud)
library(grid)
library(metR)
library(multcompView)
library(plyr)
library(png)
library(reshape2)
library(RVAideMemoire)
library(stringr)
library(vegan)
########################
## OTU CLASSIFICATION ##
########################
#Read in all data
df.controls <- read.csv("data/All_data.csv")
#Read in metadata
metadata <- read.csv("data/metadata.csv")
#Remove controls
df.all <- subset(df.controls, df.controls$Direct.Sequencing.or.Culture != "Control")
#Add Musa species
df.all$MusaSpecies <- metadata$MusaSpecies[match(substring(df.all$Serial.No, 1, 6), metadata$Serial.No)]
#Read in UNITE blastn results
unite <- read.csv("data/unite_blast_otus.tsv", sep="\t", header=FALSE, col.names=c("otu", "id", "title", "identity", "evalue", "bitscore"))
#Make empty dataframe for top UNITE hits and associated taxonomy
unite.class <- data.frame(OTU=unique(unite$otu), top.hit=NA, identity=NA, sp.=NA, gen.=NA, fam.=NA, ord.=NA, class.=NA, phy.=NA, king.=NA)
#Make vector of taxonomy levels
tax.levels <- c("sp.", "gen.", "fam.", "ord.", "class.", "phy.", "king.")
#For each OTU...
for (i in 1:length(unite.class$OTU)) {
#Pull top hit and its identity and add to dataframe
unite.class$identity[i] <- unite$identity[unite$otu == unite.class$OTU[i]][which.max(unite$identity[unite$otu == unite.class$OTU[i]])]
unite.class$top.hit[i] <- str_split(unite$title[unite$otu == unite.class$OTU[i]][which.max(unite$identity[unite$otu == unite.class$OTU[i]])], pattern="\\|")[[1]][1]
#For each taxonomy level...
for (j in 1:length(tax.levels)) {
#Add taxonomy data from UNITE
unite.class[i, tax.levels[j]] <- gsub(paste0('^.*', substring(tax.levels, 1, 1)[j], '__\\s*|\\s*;.*$'), "", str_split(unite$title[unite$otu == unite.class$OTU[i]][which.max(unite$identity[unite$otu == unite.class$OTU[i]])], pattern="\\|")[[1]][5])
}
#Replace taxonomy level classification based on following thresholds:
## >99% identity = species
## >= 98 < 99 = genus
## >= 96 < 98 = family
## >= 94 < 96 = order
## >= 92 < 94 = class
## <92 = phylum
if (unite.class$identity[i] >= 98 & unite.class$identity[i] < 99) {
unite.class$sp.[i] <- paste(gsub('^.*g__\\s*|\\s*;.*$', '', str_split(unite$title[unite$otu == unite.class$OTU[i]][which.max(unite$identity[unite$otu == unite.class$OTU[i]])], pattern="\\|")[[1]][5]), "sp.")
}
if (unite.class$identity[i] >= 96 & unite.class$identity[i] < 98) {
for (j in 1:2) {
unite.class[i, tax.levels[j]] <- paste(gsub('^.*f__\\s*|\\s*;.*$', "", str_split(unite$title[unite$otu == unite.class$OTU[i]][which.max(unite$identity[unite$otu == unite.class$OTU[i]])], pattern="\\|")[[1]][5]), tax.levels[j])
}
}
if (unite.class$identity[i] >= 94 & unite.class$identity[i] < 96) {
for (j in 1:3) {
unite.class[i, tax.levels[j]] <- paste(gsub('^.*o__\\s*|\\s*;.*$', "", str_split(unite$title[unite$otu == unite.class$OTU[i]][which.max(unite$identity[unite$otu == unite.class$OTU[i]])], pattern="\\|")[[1]][5]), tax.levels[j])
}
}
if (unite.class$identity[i] >= 92 & unite.class$identity[i] < 94) {
for (j in 1:4) {
unite.class[i, tax.levels[j]] <- paste(gsub('^.*c__\\s*|\\s*;.*$', "", str_split(unite$title[unite$otu == unite.class$OTU[i]][which.max(unite$identity[unite$otu == unite.class$OTU[i]])], pattern="\\|")[[1]][5]), tax.levels[j])
}
}
if (unite.class$identity[i] < 92) {
for (j in 1:5) {
unite.class[i, tax.levels[j]] <- paste(gsub('^.*p__\\s*|\\s*;.*$', "", str_split(unite$title[unite$otu == unite.class$OTU[i]][which.max(unite$identity[unite$otu == unite.class$OTU[i]])], pattern="\\|")[[1]][5]), tax.levels[j])
}
}
#If UNITE taxonomy contains 'unidentified'...
if (length(grep("unidentified", unite.class[i,])) > 0) {
#Correct the classification from higher levels
for (j in grep("unidentified", unite.class[i,])) {
unite.class[i, j] <- paste(unite.class[i, max(grep("unidentified", unite.class[i,])) + 1],
colnames(unite.class)[j])
}
}
}
#Remove underscores from species names
unite.class$sp. <- gsub("_", " ", unite.class$sp.)
#Make sp. endings uniform
unite.class$sp. <- sub(" sp$", " sp.", unite.class$sp.)
##Ascomycota
#Read in OTU assignments from T-BAS
tbas.class.asco <- read.csv("data/TBAS_assignments_asco_081220.csv")
#Adapt taxon level naming to match UNITE dataframe
colnames(tbas.class.asco)[c(3, 5, 7, 9, 11)] <- c("phy.", "class.", "ord.", "fam.", "gen.")
#Filter UNITE dataframe for only Ascomycota
unite.class.asco <- unite.class[unite.class$phy. == "Ascomycota",]
#Combine results for Supplementary Table
asco.table <- cbind(unite.class.asco, tbas.class.asco[match(unite.class.asco$OTU, tbas.class.asco$Query.sequence), c(13:11, 9, 7, 5, 3)])
#Add column for classification clash between UNITE and T-BAS
unite.class.asco$clash <- NA
#For each ascomycete OTU...
for (i in 1:length(unite.class.asco$OTU)) {
classified.level <- min(which(sapply(str_split(unite.class.asco[i,4:9], " "), length) == 1))
#If the assignments don't match...
if (!unite.class.asco[i, tax.levels[classified.level]] %in% tbas.class.asco[tbas.class.asco$Query.sequence == unite.class.asco$OTU[i], tax.levels[classified.level]]) {
#Add to the clash column
unite.class.asco$clash[i] <- "Y"
min.level <- min(which(!is.na(match(unite.class.asco[i,], tbas.class.asco[tbas.class.asco$Query.sequence == unite.class.asco$OTU[i],])))[-1] - 3)
for (j in 1:min.level - 1) {
unite.class.asco[i, tax.levels[j]] <- paste(gsub(paste0('^.*', substring(tax.levels[min.level], 1, 1), '__\\s*|\\s*;.*$'), "", str_split(unite$title[unite$otu == unite.class.asco$OTU[i]][which.max(unite$identity[unite$otu == unite.class.asco$OTU[i]])], pattern="\\|")[[1]][5]), tax.levels[j])
}
}
}
#Update supplementary table with clashes and consensus
asco.table <- cbind(asco.table, unite.class.asco[match(asco.table$OTU, unite.class.asco$OTU), c(11, 4:10)])
colnames(asco.table) <- c(colnames(asco.table)[1:12], "gen.tbas", "fam.tbas", "ord.tbas", "class.tbas", "phy.tbas", "clash", "sp.consensus", "gen.consensus", "fam.consensus", "ord.consensus", "class.consensus", "phy.consensus", "king.consensus")
##Non-Ascomycota
#Read in OTU assignments from T-BAS
tbas.class.nonasco <- read.csv("data/TBAS_assignments_allfungi_081220.csv")
#Filter for non-ascomycetes
tbas.class.nonasco <- tbas.class.nonasco[tbas.class.nonasco$Phylum.level.assignment != "Ascomycota",]
#Adapt taxon level naming to match UNITE dataframe
colnames(tbas.class.nonasco)[c(3, 5)] <- c("phy.", "ord.")
#Filter UNITE dataframe for non-Ascomycota
unite.class.nonasco <- unite.class[unite.class$phy. != "Ascomycota",]
#Combine results for Supplementary Table
nonasco.table <- cbind(unite.class.nonasco, tbas.class.nonasco[match(unite.class.nonasco$OTU, tbas.class.nonasco$Query.sequence), c(7:5, 3)])
#Add column for classification clash between UNITE and T-BAS
unite.class.nonasco$clash <- NA
#For each non-ascomycete OTU...
for (i in 1:length(unite.class.nonasco$OTU)) {
if (1 %in% sapply(str_split(unite.class.nonasco[i,4:9], " "), length)) {
classified.level <- min(which(sapply(str_split(unite.class.nonasco[i,4:9], " "), length) == 1))
#If the assignments don't match...
if (!unite.class.nonasco[i, tax.levels[classified.level]] %in% tbas.class.nonasco[tbas.class.nonasco$Query.sequence == unite.class.nonasco$OTU[i], tax.levels[classified.level]]) {
#Add to the clash column
unite.class.nonasco$clash[i] <- "Y"
if (length(which(!is.na(match(unite.class.nonasco[i,], tbas.class.nonasco[tbas.class.nonasco$Query.sequence == unite.class.nonasco$OTU[i],])))) > 0) {
min.level <- min(which(!is.na(match(unite.class.nonasco[i,], tbas.class.nonasco[tbas.class.nonasco$Query.sequence == unite.class.nonasco$OTU[i],])))[-1] - 3)
} else {
min.level <- 7
}
for (j in 1:min.level - 1) {
unite.class.nonasco[i, tax.levels[j]] <- paste(gsub(paste0('^.*', substring(tax.levels[min.level], 1, 1), '__\\s*|\\s*;.*$'), "", str_split(unite$title[unite$otu == unite.class.nonasco$OTU[i]][which.max(unite$identity[unite$otu == unite.class.nonasco$OTU[i]])], pattern="\\|")[[1]][5]), tax.levels[j])
}
}
}
}
classification.df <- rbind(unite.class.asco, unite.class.nonasco)
#Update supplementary table with clashes and consensus
nonasco.table <- cbind(nonasco.table, unite.class.nonasco[match(nonasco.table$OTU, unite.class.nonasco$OTU), c(11, 4:10)])
colnames(nonasco.table) <- c(colnames(nonasco.table)[1:12], "ord.tbas", "phy.tbas", "clash", "sp.consensus", "gen.consensus", "fam.consensus", "ord.consensus", "class.consensus", "phy.consensus", "king.consensus")
write.csv(rbind.fill(asco.table, nonasco.table), "Supplementary_Data_Sheet_1.csv", row.names=FALSE)
#Read in ITS OTU mapping from USEARCH
otus <- read.table("data/USEARCH_OTUS.txt", sep='\t', col.names=c("Serial.No","OTU"))
#Make main OTU dataframe
df <- df.all
#Add OTU dataframe
df["OTU"] <- otus$OTU[match(df$Serial.No, otus$Serial.No)]
#Add classification to dataframe
df[,c("Phylum","Class","Order","Genus", "Species")] <- classification.df[,c("phy.","class.","ord.","gen.", "sp.")][match(df$OTU, classification.df$OTU),]
#Remove accessions which couldn't cluster into OTUs
df <- subset(df, OTU != "#N/A")
#Remove empty rows
df <- df[!apply(is.na(df) | df == "", 1, all),]
#Remove duplicate clones
df <- subset(df, !duplicated(data.frame(Serial.No=sub("\\--.*", "", df$Serial.No), OTU=df$OTU)))
#Remove controls
df <- df[df$Direct.Sequencing.or.Culture != "Control",]
#Read in sequences flagged as mixed by GenBank
flagged <- read.csv("data/GenBank_flagged.tsv", header=FALSE)$V1
#Remove flagged sequences
df <- df[-which(!is.na(match(df$Serial.No, flagged))),]
#############################################################
## FIGURE 1 - OTUs PER SEED AND SPECIES ACCUMULATION CURVE ##
#############################################################
## FIGURE 1A - UNIQUE OTUs PER SEED ##
for (method in c("Culture", "Direct")) {
#Extract vector of seeds with OTUs
seed <- substring(df$Serial.No[df$Direct.Sequencing.or.Culture == method],1,9)[!duplicated(substring(df$Serial.No[df$Direct.Sequencing.or.Culture == method], 1, 9))]
seed.df <- data.frame(Serial.No=seed, count=NA)
#Count number of fungi per seed (all Musa accessions)
for (i in 1:length(seed)) {
seed.df$count[i] <- length(grep(seed[i], df$Serial.No))
}
seed.count <- data.frame(x=sort(unique(seed.df$count)), y=NA)
for (i in 1:length(seed.count$x)) {
seed.count$y[i] <- length(seed.df$count[seed.df$count == seed.count$x[i]])
}
#Count number of fungi per seed (by Musa species)
#Add Musa species column to seed dataframe
seed.df["musa"] <- metadata$MusaSpecies[match(gsub("-.*", "", seed.df$Serial.No), metadata$Serial.No)]
#To species level
seed.df$musa <- word(seed.df$musa, 1, 2)
#Make vector of Musa species
musa <- sort(unique(word(unique(seed.df$musa), 1, 2)))
#Make dataframe to add counts to
seed.count <- data.frame(x=sort(unique(seed.df$count)), Macu=NA, Mbal=NA, Miti=NA, Mvel=NA, Mvio=NA, method=method)
for (i in 1:length(musa)) {
for (j in 1:length(seed.count$x)) {
seed.count[j, i+1] <- length(seed.df$count[seed.df$count == seed.count$x[j] & seed.df$musa == musa[i]])
}
}
assign(paste0("seed.count.", method), seed.count)
assign(paste0("seed.df.", method), seed.df)
}
#Combine culture and direct seed counts
seed.count <- rbind(seed.count.Culture, seed.count.Direct)
#Create dataframe for total number of seeds per species
df.tot <- data.frame(musa=rep(musa, each=2), x=0, y=NA, method=rep(c("Culture", "Direct"), times=length(musa)))
#Add total seed count
for (i in df.tot$musa) {
df.tot[df.tot$method == "Culture" & df.tot$musa == i, "y"] <- length(unique(substring(df.all$Serial.No[grep(i, df.all$MusaSpecies)][df.all$Direct.Sequencing.or.Culture[grep(i, df.all$MusaSpecies)] == "Culture"], 1, 9)))
df.tot[df.tot$method == "Direct" & df.tot$musa == i, "y"] <- length(unique(substring(df.all$Serial.No[grep(i, df.all$MusaSpecies)][df.all$Direct.Sequencing.or.Culture[grep(i, df.all$MusaSpecies)] == "Direct"], 1, 9)))
}
#Remove numbers of seeds with isolates from the total seed count
df.tot$y[df.tot$method == "Culture"] <- df.tot$y[df.tot$method == "Culture"] - colSums(seed.count.Culture[2:6])
df.tot$y[df.tot$method == "Direct"] <- df.tot$y[df.tot$method == "Direct"] - colSums(seed.count.Direct[2:6])
#Combine counts for all accessions and for separate Musa species into one dataframe for plotting
seed.df <- rbind(df.tot,
data.frame(x=seed.count$x, y=seed.count$Macu, musa=musa[1], method=seed.count$method),
data.frame(x=seed.count$x, y=seed.count$Mbal, musa=musa[2], method=seed.count$method),
data.frame(x=seed.count$x, y=seed.count$Miti, musa=musa[3], method=seed.count$method),
data.frame(x=seed.count$x, y=seed.count$Mvel, musa=musa[4], method=seed.count$method),
data.frame(x=seed.count$x, y=seed.count$Mvio, musa=musa[5], method=seed.count$method),
data.frame(x=0, y=rep(20, 2), musa="Musa gracilis", method=c("Culture", "Direct")))
#Custom order of Musa species in plot
seed.df$musa <- factor(seed.df$musa, levels=c("Musa acuminata", "Musa balbisiana", "Musa itinerans", "Musa velutina", "Musa violascens", "Musa gracilis"))
#Replace labels
seed.df$method <- sub("Culture", "Culture-dependent", seed.df$method)
seed.df$method <- sub("Direct", "Culture-independent", seed.df$method)
#Plot seed count bargraphs
gg.seed <- ggplot(seed.df, aes(x=x, y=y)) +
geom_bar(stat="identity") +
geom_text(data=subset(seed.df, y > 0), aes(label=y), nudge_y=15, size=2) +
facet_grid(method ~ musa) +
scale_x_discrete(name="Number of unique OTUs",
limits=c(seq(0,7,1))) +
scale_y_continuous(name="Number of seeds",
limits=c(0, 256),
breaks=seq(0, 255, 50),
expand=c(0, 5)) +
theme_minimal() +
theme(axis.title.x=element_text(margin=unit(c(3, 0, 0, 0), "mm")),
axis.title.y=element_text(margin=unit(c(0, 3, 0, 0), "mm")),
panel.spacing.y=unit(1, "lines"),
strip.text.x=element_text(face="italic", size=7),
strip.text.y=element_text(size=9),
plot.title.position="plot") +
labs(subtitle=expression(bold("A")))
## FIGURE 1B - SPECIES ACCUMULATION CURVE ##
#Make dataframe of species
spec.df <- metadata[c("Serial.No", "MusaSpecies")]
#Create OTU abundance matrix
otu.count <- matrix(nrow=length(unique(spec.df$Serial.No)),ncol=length(unique(df$OTU)))
colnames(otu.count) <- unique(df$OTU)
rownames(otu.count) <- unique(spec.df$Serial.No)
for (i in 1:length(rownames(otu.count))) {
for (j in 1:length(colnames(otu.count))) {
otu.count[i,j] <- length(df$OTU[grep(rownames(otu.count)[i], df$Serial.No)][df$OTU[grep(rownames(otu.count)[i], df$Serial.No)] == colnames(otu.count)[j]])
}
}
#Accumulation curves including singletons
#Create dataframe to collect rarefaction accumulation curve data for acuminata, balbisians and itinerans
spec.accum.df <- data.frame(individual=NA, richness=NA, error=NA, musa=sort(spec.df$MusaSpecies[spec.df$MusaSpecies == musa[1] | spec.df$MusaSpecies == musa[2] | spec.df$MusaSpecies == musa[3]]), rare="Including singletons")
#Rarefaction for acuminata, balbisians and itinerans
for (i in 1:length(musa[1:3])) {
#Subset OTU abundance matrix for top three host species
tmp.df <- subset(otu.count, rownames(otu.count) %in% rownames(otu.count)[spec.df$MusaSpecies[match(rownames(otu.count), spec.df$Serial.No)] == musa[i]])
#Species accumulation
tmp.accum <- specaccum(tmp.df, method="rarefaction")
#Add to results dataframe
spec.accum.df[spec.accum.df$musa == musa[i],] <- data.frame(individual=tmp.accum$sites, richness=tmp.accum$richness, error=tmp.accum$sd, musa=musa[i], rare="Including singletons")
}
#Rarefaction for all Musa accessions
tmp.accum <- specaccum(otu.count, method="rarefaction")
tmp.accum.df <- data.frame(individual=tmp.accum$sites, richness=tmp.accum$richness, error=tmp.accum$sd, musa="All", rare="Including singletons")
#Combine species accumulation data including singletons
spec.accum.df <- rbind(spec.accum.df, tmp.accum.df)
#Accumulation curves excluding singletons
#Create dataframe to collect rarefaction accumulation curve data for acuminata, balbisians and itinerans
spec.accum.excl.df <- data.frame(individual=NA, richness=NA, error=NA, musa=sort(spec.df$MusaSpecies[spec.df$MusaSpecies == musa[1] | spec.df$MusaSpecies == musa[2] | spec.df$MusaSpecies == musa[3]]), rare="Excluding singletons")
#Remove singleton OTUs
otu.count.excl <- subset(otu.count, select=-which(colSums(otu.count) == 1))
#Rarefaction for acuminata, balbisians and itinerans
for (i in 1:length(musa[1:3])) {
#Subset for top three host species
tmp.df <- subset(otu.count.excl, rownames(otu.count.excl) %in% rownames(otu.count.excl)[spec.df$MusaSpecies[match(rownames(otu.count.excl), spec.df$Serial.No)] == musa[i]])
#Species accumulation
tmp.accum <- specaccum(tmp.df, method="rarefaction")
#Add to results dataframe
spec.accum.excl.df[spec.accum.excl.df$musa == musa[i],] <- data.frame(individual=tmp.accum$sites, richness=tmp.accum$richness, error=tmp.accum$sd, musa=musa[i], rare="Excluding singletons")
}
#Rarefaction for all Musa accessions
tmp.accum <- specaccum(otu.count.excl, method="rarefaction")
tmp.accum.df <- data.frame(individual=tmp.accum$sites, richness=tmp.accum$richness, error=tmp.accum$sd, musa="All", rare="Excluding singletons")
#Combine species accumulation data including and excluding singletons
spec.accum.df <- rbind(spec.accum.excl.df, tmp.accum.df, spec.accum.df)
#Plot accumulation curves
gg.accum <- ggplot(spec.accum.df, aes(x=individual, y=richness, color=musa)) +
facet_wrap(. ~ rare, scales="free", nrow=1) +
geom_ribbon(aes(ymax=richness+error, ymin=richness-error), linetype=0, fill="grey", alpha=0.4, show.legend=FALSE) +
geom_line(size=0.5) +
scale_color_manual(values=c("black", "#66c2a5","#fc8d62","#8da0cb"),
name=expression(paste(italic("Musa")," species")),
labels=c("All", expression(italic("Musa acuminata")), expression(italic("Musa balbisiana")), expression(italic("Musa itinerans")))) +
labs(x=expression(paste("Number of ",italic("Musa")," accessions sampled")), y="Number of unique OTUs") +
theme_minimal() +
theme(axis.title.x=element_text(margin=unit(c(3, 0, 0, 0), "mm")),
axis.title.y=element_text(margin=unit(c(0, 3, 0, 0), "mm")),
legend.text.align=0,
plot.title.position="plot") +
labs(subtitle=expression(bold("B")))
#Write to file - FIGURE 1
tiff(file=paste0("Figure_1-", Sys.Date(), ".tiff"), height=18, width=15, units="cm", res=300, compression="lzw")
ggpubr::ggarrange(gg.seed, gg.accum, nrow=2, heights=c(1.5,1))
dev.off()
###################################
## FIGURE 2 - OTU CLASSIFICATION ##
###################################
## T-BAS VISUALISATION OF OTU ASSIGNMENT ##
#Make dataframe of OTU count per isolation method
method.count <- data.frame(otu=unique(df$OTU), direct=NA, culture=NA, clone=NA)
for (i in 1:length(unique(df$OTU))) {
method.count$direct[i] <- sum(df$Direct.Sequencing.or.Culture[df$OTU == unique(df$OTU)[i]] == "Direct" & df$Clone[df$OTU == unique(df$OTU)[i]] != "Y")
method.count$culture[i] <- sum(df$Direct.Sequencing.or.Culture[df$OTU == unique(df$OTU)[i]] == "Culture")
method.count$clone[i] <- sum(df$Clone[df$OTU == unique(df$OTU)[i]] == "Y")
}
#Read in T-BAS tree
tbas <- read.tree("data/TBAS_asco_081220.tre")
#Make dataframe of isolation method
tbas.df <- data.frame(tip=tbas$tip.label, direct=NA, culture=NA, clone=NA)
#Add counts to T-BAS plot dataframe
tbas.df$direct <- method.count$direct[match(tbas.df$tip, method.count$otu)]
tbas.df$culture <- method.count$culture[match(tbas.df$tip, method.count$otu)]
tbas.df$clone <- method.count$clone[match(tbas.df$tip, method.count$otu)]
#Replace 0 with NA
tbas.df[tbas.df == 0] <- NA
#Read in node and colour data for T-BAS tree
cladelabels <- read.csv("data/TBAS_clades.csv")
#Plot tree
gg.tbas <- ggtree(tbas,
layout="circular",
cex=0.2,
branch.length="none")
for (i in which(!is.na(cladelabels$node))) {
gg.tbas <- gg.tbas +
geom_hilight(node=cladelabels[i,2], fill=cladelabels[i,3])
}
gg.tbas <- gg.tbas %<+% tbas.df +
geom_tree(cex=0.2) +
geom_tippoint(aes(x=x+1, size=culture, colour="Culturing"), na.rm=TRUE) +
geom_tippoint(aes(x=x+3, size=direct, colour="Direct extraction"), na.rm=TRUE) +
geom_tippoint(aes(x=x+3, size=clone, colour="Cloning"), alpha=0.4, na.rm=TRUE) +
scale_colour_manual(name="Sampling method", labels=c("Cloning", "Culturing","Direct sequencing"), values=c("black", "#E69F00", "#56B4E9")) +
labs(size="OTU abundance") +
guides(colour=guide_legend(override.aes=list(size=3))) +
theme(legend.position="right",
plot.margin=unit(c(-10,0,-10,0), "mm"),
legend.box.margin=margin(c(0,0,0,0)),
plot.title.position="plot")
## PIECHART VISUALISATION OF OTU CLASSIFICATION ##
#Make taxonomy dataframe
tmp <- data.frame(phylum=df$Phylum, class=df$Class, order=df$Order, genus=df$Genus, species=df$Species)
#Filter for Ascomycota
pie.df <- tmp[tmp$phylum == "Ascomycota",]
#Remove duplicate rows for genera
pie.df <- pie.df[!duplicated(pie.df$species),]
#Add column of total count for each genus
pie.df["num"] <- NA
for (i in 1:length(pie.df$species)) {
pie.df$num[i] <- sum(df$Species == pie.df$species[i])
}
#Sort dataframe sequentially by class, order and genus
pie.df <- transform(pie.df, freq=ave(seq(nrow(pie.df)), class, FUN=length))
pie.df <- transform(pie.df, freq2=ave(seq(nrow(pie.df)), order, FUN=length))
pie.df <- transform(pie.df, freq3=ave(seq(nrow(pie.df)), genus, FUN=length))
pie.df <- pie.df[order(-pie.df["freq"], -pie.df["freq2"], -pie.df["freq3"], -pie.df["num"], pie.df["species"]), ]
#Add ymin and ymax columns for size of pie slice
ymin <- c(0,pie.df$num)
for (i in 2:length(ymin)) {
ymin[i] <- ymin[i] + ymin[i-1]
}
pie.df["ymin"] <- ymin[1:length(pie.df$species)]
pie.df["ymax"] <- ymin[2:length(ymin)]
#Add column with midpoint position for genus label in pie slice
pie.df["lab.pos.sp"] <- (pie.df$ymax + pie.df$ymin) / 2
for (i in c("genus", "order", "class")) {
#Make dataframe of genus label positions
pos <- data.frame(taxon=unique(pie.df[,i]), lab.pos=NA)
for (j in 1:length(unique(pie.df[,i]))){
pos$lab.pos[pos$taxon == unique(pie.df[,i])[j]] <- pie.df$ymax[pie.df[,i] == unique(pie.df[,i])[j]][which.max(pie.df$ymax[pie.df[,i] == unique(pie.df[,i])[j]])]
}
ymin <- c(0, pos$lab.pos)
pos["ymin"] <- ymin[1:length(pos$taxon)]
pos["ymax"] <- ymin[2:length(ymin)]
res <- vector(length=length(pos$lab.pos))
for (k in 2:length(ymin)) {
res[k-1] <- (ymin[k] + ymin[k-1]) / 2
}
pos$lab.pos <- res
assign(paste0("pos.", i), pos)
}
#Add unclassified Ascomycota to the T-BAS clade colours
colours <- rbind(cladelabels[-2], data.frame(class="Ascomycota class.", colour="#818181", mid="#9b9b9b", low="#b5b5b5", lower="#b5b5b5", lowest="#c7c7c7"))
#Sort the T-BAS clade colours according to the count data
colours <- colours[match(unique(pie.df$class), colours$class),]
#Add colour columns to main dataframe
for (i in 1:length(unique(pie.df$class))) {
#Class
pie.df$colour.class[pie.df$class == unique(pie.df$class)[i]] <- as.character(colours$colour[i])
#Order
tmp <- subset(pie.df, class == unique(pie.df$class)[i])
for (j in 1:length(unique(pie.df$order[pie.df$class == unique(pie.df$class)[i]]))) {
pie.df$colour.order[pie.df$order == unique(tmp$order)[j]] <- colorRampPalette(c(as.character(colours$mid[i]),as.character(colours$low[i])))(length(unique(pie.df$order[pie.df$class == unique(pie.df$class)[i]])))[j]
}
tmp <- subset(pie.df, class == unique(pie.df$class)[i])
for (j in 1:length(unique(pie.df$genus[pie.df$class == unique(pie.df$class)[i]]))) {
pie.df$colour.genus[pie.df$genus == unique(tmp$genus)[j]] <- colorRampPalette(c(as.character(colours$low[i]),as.character(colours$lower[i])))(length(unique(pie.df$genus[pie.df$class == unique(pie.df$class)[i]])))[j]
}
#Species
pie.df$colour.species[pie.df$class == unique(pie.df$class)[i]] <- colorRampPalette(c(as.character(colours$lower[i]),as.character(colours$lowest[i])))(length(unique(pie.df$species[pie.df$class == unique(pie.df$class)[i]])))
}
#Create a vector with names from all taxon levels and sort alphabetically
all.tax <- c(as.character(pie.df$class), as.character(pie.df$order), as.character(pie.df$genus), as.character(pie.df$species))
all.tax <- all.tax[order(all.tax)]
all.tax <- all.tax[!duplicated(all.tax)]
#Create a dataframe of colours for ggplot
pie.colours <- rbind(data.frame(name=pie.df$class[!duplicated(pie.df$class)],
colour=pie.df$colour.class[!duplicated(pie.df$class)]),
data.frame(name=pie.df$order[!duplicated(pie.df$order)],
colour=pie.df$colour.order[!duplicated(pie.df$order)]),
data.frame(name=pie.df$genus[!duplicated(pie.df$genus)],
colour=pie.df$colour.genus[!duplicated(pie.df$genus)]),
data.frame(name=pie.df$species[!duplicated(pie.df$species)],
colour=pie.df$colour.species[!duplicated(pie.df$species)]))
#Sorted alphabetically and by taxonomic level
pie.colours <- pie.colours[order(pie.colours$name),]
#Make into vector
pie.colours <- as.vector(pie.colours$colour[match(all.tax, pie.colours$name)])
#Edit Fusarium solani name
pie.df$species[grep("Fusarium solani", pie.df$species)] <- "Fusarium solani \u2020"
set.seed(1)
#Plot piechart
gg.pie <- ggplot(pie.df) +
geom_rect(aes(fill=species, ymax=ymax, ymin=ymin, xmax=6, xmin=4.5),
colour="white",
show.legend=FALSE) +
geom_rect(data=pos.genus,
aes(fill=taxon, ymax=ymax, ymin=ymin, xmax=4.5, xmin=3),
colour="white",
show.legend=FALSE) +
geom_rect(data=pos.order,
aes(fill=taxon, ymax=ymax, ymin=ymin, xmax=3, xmin=1.5),
colour="white",
show.legend=FALSE) +
geom_rect(data=pos.class,
aes(fill=taxon, ymax=ymax, ymin=ymin, xmax=1.5, xmin=0),
colour="white",
show.legend=FALSE) +
annotate("text", x=0.75, y=max(pie.df$ymax)*0.75,
label="Class", color="white", size=3, fontface="bold") +
annotate("text", x=2.25, y=max(pie.df$ymax)*0.75,
label="Order", color="white", size=3, fontface="bold") +
annotate("text", x=3.75, y=max(pie.df$ymax)*0.75,
label="Genus", color="white", size=3, fontface="bold") +
annotate("text", x=5.25, y=max(pie.df$ymax)*0.75,
label="Species", color="white", size=3, fontface="bold") +
geom_label_repel(aes(x=6, y=lab.pos.sp, label=species, fill=species),
#direction="y",
fontface="italic",
size=2,
nudge_x=1,
show.legend=FALSE) +
geom_label_repel(data=pos.order[-c(grep("ord", pos.order$taxon), 15),],
aes(x=2.25, y=lab.pos, label=taxon, fill=taxon),
direction="y",
size=2,
nudge_x=0.5,
show.legend=FALSE) +
geom_point(data=pos.class,
aes(x=1, y=lab.pos, color=taxon),
size=2,
shape=15,
alpha=0) +
scale_fill_manual(values=pie.colours) +
scale_color_manual(name="", labels=c("Unclassified Ascomycota",as.character(sort(pos.class$taxon)[-1])), values=as.vector(colours$colour[order(as.vector(colours$class))])) +
xlim(c(0, 8)) +
coord_polar(theta="y") +
theme_void() +
guides(color=guide_legend(override.aes=list(alpha=1, size=5))) +
theme(legend.position="left",
plot.margin=unit(c(-10,0,-10,0), "mm"),
legend.box.margin=margin(c(0,-80,0,0)))
#Write to file - FIGURE 2
tiff(file=paste0("Figure_2-", Sys.Date(), ".tiff"), height=26, width=20, units="cm", res=300, compression="lzw")
ggpubr::ggarrange(gg.tbas, gg.pie, heights=c(1,1.5), ncol=1)
dev.off()
################################################
## FIGURE 3 - COMPARISON OF DETECTION METHODS ##
################################################
#Test for difference in OTUs between methods
#Create OTU abundance matrix with accessions split into three methods
otu.count.methods <- matrix(nrow=length(unique(spec.df$Serial.No))*3,ncol=length(unique(df$OTU)))
colnames(otu.count.methods) <- unique(df$OTU)
rownames(otu.count.methods) <- c(paste0(unique(spec.df$Serial.No), ".culture"),
paste0(unique(spec.df$Serial.No), ".direct"),
paste0(unique(spec.df$Serial.No), ".clone"))
#For each row...
for (i in 1:length(rownames(otu.count.methods))) {
#For each OTU...
for (j in 1:length(colnames(otu.count.methods))) {
#Grab the accession
accession <- sub("\\..*", "", rownames(otu.count.methods)[i])
#If the method is culturing, count the number of occurrences of the OTU and enter into matrix
if (sub(".*\\.", "", rownames(otu.count.methods)[i]) == "culture") {
otu.count.methods[i,j] <- length(df$OTU[grep(accession, df$Serial.No)[which(grepl(accession, df$Serial.No)) %in% which(df$Direct.Sequencing.or.Culture == "Culture")]][df$OTU[grep(accession, df$Serial.No)[which(grepl(accession, df$Serial.No)) %in% which(df$Direct.Sequencing.or.Culture == "Culture")]] == colnames(otu.count.methods)[j]])
}
#If the method is direct sequencing, count the number of occurrences of the OTU and enter into matrix
if (sub(".*\\.", "", rownames(otu.count.methods)[i]) == "direct") {
otu.count.methods[i,j] <- length(df$OTU[grep(accession, df$Serial.No)[which(grepl(accession, df$Serial.No)) %in% which(df$Direct.Sequencing.or.Culture == "Direct" & df$Clone != "Y")]][df$OTU[grep(accession, df$Serial.No)[which(grepl(accession, df$Serial.No)) %in% which(df$Direct.Sequencing.or.Culture == "Direct" & df$Clone != "Y")]] == colnames(otu.count.methods)[j]])
}
#If the method is cloning, count the number of occurrences of the OTU and enter into matrix
if (sub(".*\\.", "", rownames(otu.count.methods)[i]) == "clone") {
otu.count.methods[i,j] <- length(df$OTU[grep(accession, df$Serial.No)[which(grepl(accession, df$Serial.No)) %in% which(df$Clone == "Y")]][df$OTU[grep(accession, df$Serial.No)[which(grepl(accession, df$Serial.No)) %in% which(df$Clone == "Y")]] == colnames(otu.count.methods)[j]])
}
}
}
#Remove rows with no occurrences
otu.count.methods <- otu.count.methods[-which(rowSums(otu.count.methods) == 0),]
#betadisper by methods
anova(betadisper(vegdist(otu.count.methods, method="bray"), sub(".*\\.", "", rownames(otu.count.methods))))
#ANOSIM by methods
method.anosim <- anosim(otu.count.methods, sub(".*\\.", "", rownames(otu.count.methods)), distance="bray", permutations=999)
#Indices for OTUs found by different combinations of detection methods
#Direct
direct <- which(method.count$direct > 0 & method.count$culture == 0 & method.count$clone == 0)
#Culture
culture <- which(method.count$direct == 0 & method.count$culture > 0 & method.count$clone == 0)
#Clone
clone <- which(method.count$direct == 0 & method.count$culture == 0 & method.count$clone > 0)
#Direct&Culture
direct.culture <- which(method.count$direct > 0 & method.count$culture > 0 & method.count$clone == 0)
#Direct&Clone
direct.clone <- which(method.count$direct > 0 & method.count$culture == 0 & method.count$clone > 0)
#Culture&Clone
culture.clone <- which(method.count$direct == 0 & method.count$culture > 0 & method.count$clone > 0)
#All
direct.culture.clone <- which(method.count$direct > 0 & method.count$culture > 0 & method.count$clone > 0)
#Make matrix of number of OTUs for each detection method
method.otus <- c("Direct"=length(method.count$otu[direct]),
"Culture"=length(method.count$otu[culture]),
"Clone"=length(method.count$otu[clone]),
"Direct&Culture"=length(method.count$otu[direct.culture]),
"Direct&Clone"=length(method.count$otu[direct.clone]),
"Culture&Clone"=length(method.count$otu[culture.clone]),
"Direct&Culture&Clone"=length(method.count$otu[direct.culture.clone]))
#Create and plot euler diagram
set.seed(1)
method.euler <- euler(method.otus, shape="ellipse")
euler <- plot(method.euler,
fills=list(fill=c("#56B4E9", "#E69F00", "black"), alpha=0.3),
labels=FALSE,
#quantities=list(col="white"),
edges=FALSE)
#Convert to ggplot
gg.euler <- as.ggplot(euler)
#Create dataframe for OTU wordcloud labels
euler.df <- rbind(data.frame(otu=df$Species[match(method.count$otu[direct], df$OTU)],
count=rowSums(method.count[direct, 2:4]),
method="direct",
x=0.3, y=0.25),
data.frame(otu=df$Species[match(method.count$otu[culture], df$OTU)],
count=rowSums(method.count[culture, 2:4]),
method="culture",
x=0.7, y=0.25),
data.frame(otu=df$Species[match(method.count$otu[clone], df$OTU)],
count=rowSums(method.count[clone, 2:4]),
method="clone",
x=0.52, y=0.8),
data.frame(otu=df$Species[match(method.count$otu[direct.culture], df$OTU)],
count=rowSums(method.count[direct.culture, 2:4]),
method="direct.culture",
x=0.52, y=0.34),
data.frame(otu=df$Species[match(method.count$otu[direct.clone], df$OTU)],
count=rowSums(method.count[direct.clone, 2:4]),
method="direct.clone",
x=0.42, y=0.59),
data.frame(otu=df$Species[match(method.count$otu[culture.clone], df$OTU)],
count=rowSums(method.count[culture.clone, 2:4]),
method="culture.clone",
x=0.63, y=0.51),
data.frame(otu=df$Species[match(method.count$otu[direct.culture.clone], df$OTU)],
count=rowSums(method.count[direct.culture.clone, 2:4]),
method="direct.culture.clone",
x=0.5, y=0.49))
#Plot euler
gg.euler <- gg.euler +
geom_text(x=0.1, y=0.5, size=4, fontface="bold", colour="#56B4E9",
label=paste0("Direct sequencing\n", length(method.count$otu[direct]))) +
geom_text(x=0.85, y=0.5, colour="#E69F00", size=4, fontface="bold",
label=paste0("Culture\n", length(method.count$otu[culture]))) +
geom_text(x=0.3, y=0.9, colour="black", size=4, fontface="bold",
label=paste0("Clone\n", length(method.count$otu[clone]))) +
geom_text(x=0.38, y=0.5, colour="white",
size=5, fontface="bold",
label=length(method.count$otu[direct.clone])) +
geom_text(x=0.515, y=0.285, colour="white",
size=5, fontface="bold",
label=length(method.count$otu[direct.culture])) +
geom_text(x=0.6, y=0.58, colour="white",
size=5, fontface="bold",
label=length(method.count$otu[culture.clone])) +
geom_text(x=0.5, y=0.59, colour="white",
size=5, fontface="bold",
label=length(method.count$otu[direct.culture.clone])) +
geom_text(x=0.85, y=0.9, size=3, fontface="bold",
label=paste0("ANOSIM")) +
geom_text(x=0.85, y=0.85, size=3,
label=paste0("R=", round(method.anosim$statistic, 2), " p=", round(method.anosim$signif, 3))) +
geom_text_wordcloud(data=euler.df,
eccentricity=1,
area_corr_power=1,
aes(label=otu, colour=method, size=count)) +
scale_size_continuous(range=c(1, 3)) +
scale_colour_manual(values=c("#000000", #clone
"#E69F00", #culture
hex(mixcolor(alpha=0.5, hex2RGB("#E69F00"), hex2RGB("#000000"))), #culture.clone
"#56B4E9", #direct
hex(mixcolor(alpha=0.5, hex2RGB("#56B4E9"), hex2RGB("#000000"))), #direct.clone
hex(mixcolor(alpha=0.5, hex2RGB("#56B4E9"), hex2RGB("#E69F00"))), #direct.culture
hex(mixcolor(alpha=0.5, (mixcolor(alpha=0.5, hex2RGB("#E69F00"), hex2RGB("#000000"))), hex2RGB("#56B4E9"))))) + #direct.culture.clone
theme(legend.position="none")
#Write to file - FIGURE 3
tiff(file=paste0("Figure_3-", Sys.Date(), ".tiff"), height=12, width=18, units="cm", res=300, compression="lzw")
gg.euler
dev.off()
################################################################
## FIGURE 4 - COMMUNITY COMPOSITION, DIVERSITY AND ABUNDANCE ##
################################################################
## FIGURE 4A - NMDS ANALYSIS ##
#Subset OTU abundance matrix by Musa accessions for acuminata, balbisiana and itinerans
otu.count.nmds <- subset(otu.count, rownames(otu.count) %in% rownames(otu.count)[spec.df$MusaSpecies[match(rownames(otu.count), spec.df$Serial.No)] == musa[1] | spec.df$MusaSpecies[match(rownames(otu.count), spec.df$Serial.No)] == musa[2] | spec.df$MusaSpecies[match(rownames(otu.count), spec.df$Serial.No)] == musa[3]])
#Filter for most abundant OTUs
otu.count.nmds <- otu.count.nmds[,-which(colSums(otu.count.nmds) < 20)]
#Remove rows with no OTUs
otu.count.nmds <- otu.count.nmds[-which(rowSums(otu.count.nmds) == 0),]
#Function to perform NMDS for 1-10 dimensions to pick optimal number of dimensions
NMDS.scree <- function(x) {
plot(rep(1, 10), replicate(10, metaMDS(x, autotransform=F, k=1)$stress), xlim=c(1, 10),ylim=c(0, 0.30), xlab="# of Dimensions", ylab="Stress")
for (i in 1:10) {
points(rep(i + 1,10),replicate(10, metaMDS(x, autotransform=F, k=i + 1)$stress))
}
}
#Write to file scree and stress plot - SUPPLEMENTARY FIGURE 1
tiff(file=paste0("Supplementary_Figure_1-", Sys.Date(), ".tiff"), height=15, width=12, units="cm", res=300, compression="lzw")
#Plot screeplot
par(mfrow=c(2,1), mar=c(4,4,2,2))
NMDS.scree(otu.count.nmds)
set.seed(2)
#NMDS with optimal dimensions from scree plot
NMDS <- metaMDS(otu.count.nmds, k=6, trymax=1000, trace=F)
#Plot stressplot
stressplot(NMDS)
dev.off()
#Make dataframes of NMDS site scores for ggplot (site=Musa accession)
metadata.nmds <- as.data.frame(scores(NMDS))
metadata.nmds$Serial.No <- rownames(metadata.nmds)
#Add Musa species
metadata.nmds$MusaSpecies <- spec.df$MusaSpecies[match(metadata.nmds$Serial.No, spec.df$Serial.No)]
#Add habitat
metadata.nmds$Habitat.summary <- metadata$Habitat.summary[match(metadata.nmds$Serial.No, metadata$Serial.No)]
#Add TZ
metadata.nmds$TZ <- metadata$TZ[match(metadata.nmds$Serial.No, metadata$Serial.No)]
#Add germination rate
metadata.nmds$Germination <- metadata$ER.shoots[match(metadata.nmds$Serial.No, metadata$Serial.No)] / metadata$ER.embryos[match(metadata.nmds$Serial.No, metadata$Serial.No)] * 100
#Make dataframe of NMDS species scores for ggplot (species=fungal OTUs)
otu.gg <- as.data.frame(scores(NMDS, "species"))
otu.gg$label <- df$Species[match(rownames(otu.gg), df$OTU)]
#Edit Fusarium solani name
otu.gg$label[grep("Fusarium solani", otu.gg$label)] <- "Fusarium solani \u2020"
#Fit TZ variable to NMDS
nmds.ordi.tz <- ordisurf(NMDS ~ metadata.nmds$TZ, plot=FALSE)
#Pull out coordinates for plotting
nmds.ordi.tz.gg <- expand.grid(x=nmds.ordi.tz$grid$x, y=nmds.ordi.tz$grid$y)
#Get z scores
nmds.ordi.tz.gg$z <- as.vector(nmds.ordi.tz$grid$z)
#Remova NAs
nmds.ordi.tz.gg <- data.frame(na.omit(nmds.ordi.tz.gg))
#Fit germination rate variable to NMDS
nmds.ordi.germ <- ordisurf(NMDS ~ metadata.nmds$Germination, plot=FALSE)
#Pull out coordinates for plotting
nmds.ordi.germ.gg <- expand.grid(x=nmds.ordi.germ$grid$x, y=nmds.ordi.germ$grid$y)
#Get z scores
nmds.ordi.germ.gg$z <- as.vector(nmds.ordi.germ$grid$z)
#Remova NAs
nmds.ordi.germ.gg <- data.frame(na.omit(nmds.ordi.germ.gg))
##PERMANOVA
#PERMANOVA on accessions with subset of common taxa used in NMDS
#Marginal PERMANOVA to test for unique variable impact
set.seed(1)
nmds.permanova.marginal <- as.data.frame(adonis2(otu.count.nmds ~ metadata.nmds$Habitat.summary + metadata.nmds$MusaSpecies, method="bray", perm=999, by="margin"))
#Main PERMANOVA
set.seed(1)
nmds.permanova <- as.data.frame(adonis(otu.count.nmds ~ metadata.nmds$Habitat.summary + metadata.nmds$MusaSpecies, method="bray", perm=999)$aov.tab)
#PERMANOVA on all accessions including rare taxa (excluding oil palm plantation and Kew accessions)
#Filter OTU counts for non-empty accessions
otu.count.rare <- otu.count[-which(rowSums(otu.count) == 0),]
#Exclude low-sampled habitat accessions
otu.count.rare <- subset(otu.count.rare, rownames(otu.count.rare) %in% metadata$Serial.No[-c(35, 36, 39, 45)])
#Filter metadata similarly
metadata.rare <- metadata[match(rownames(otu.count.rare), metadata$Serial.No),]
#Marginal PERMANOVA to test for unique variable impact
set.seed(1)
permanova.marginal <- as.data.frame(adonis2(otu.count.rare ~ metadata.rare$Habitat.summary + metadata.rare$MusaSpecies, method="bray", perm=999, by="margin"))
#Main PERMANOVA
set.seed(1)
permanova <- as.data.frame(adonis(otu.count.rare ~ metadata.rare$Habitat.summary + metadata.rare$MusaSpecies, method="bray", perm=999)$aov.tab)
#Combine PERMANOVA testing into dataframe
Table.1 <- merge(rbind(data.frame(variable=rownames(nmds.permanova.marginal[1:2,]),
nmds.permanova.marginal[1:2, c(3,5)]),
data.frame(variable=rownames(permanova.marginal[1:2,]),
permanova.marginal[1:2, c(3,5)])),
rbind(data.frame(variable=rownames(nmds.permanova[1:2,]),
nmds.permanova[1:2, c(5,6)]),
data.frame(variable=rownames(permanova[1:2,]),
permanova[1:2, c(5,6)])), by="variable", all=TRUE)
#PERMDISP
#For each of the data subsets (common and all)...
for (i in c("nmds", "rare")) {
#For each variable...
for (j in c("Habitat.summary", "MusaSpecies")) {
#betadisper analysis
disp <- betadisper(vegdist(get(paste0("otu.count.", i)), method="bray"), get(paste0("metadata.", i))[,j])
#ANOVA for significance
disp.anova <- anova(betadisper(vegdist(get(paste0("otu.count.", i)), method="bray"),
get(paste0("metadata.", i))[,j]))
#Print results based on p-value significance
if (disp.anova$`Pr(>F)`[1] >= 0.05) {
print(paste(i, j, ": even data dispersion, p=", signif(disp.anova$`Pr(>F)`[1], digits=3)))
} else {
print(paste(i, j, ": uneven data dispersion, p=", signif(disp.anova$`Pr(>F)`[1], digits=3)))
}
assign(paste0("disp.anova.", i, ".", j), disp.anova)
#For the habitat variable...
if (j == "Habitat.summary") {
#Make a dataframe with the dispersion distances from centroids for each accession
disp.df <- as.data.frame(disp$distances)
colnames(disp.df) <- "distances"
#Add accession
disp.df$accession <- rownames(disp.df)
#Add habitat
disp.df$hab <- metadata$Habitat.summary[match(disp.df$accession, metadata$Serial.No)]
#Add data subset
disp.df$data <- i
assign(paste0("disp.", i, ".", j), disp.df)
}
}
}
#Add PERMDISP results to table
Table.1$betadisper.p <- NA
Table.1$betadisper.p[1] <- disp.anova.nmds.Habitat.summary$`Pr(>F)`[1]
Table.1$betadisper.p[2] <- disp.anova.nmds.MusaSpecies$`Pr(>F)`[1]
Table.1$betadisper.p[3] <- disp.anova.rare.Habitat.summary$`Pr(>F)`[1]
Table.1$betadisper.p[4] <- disp.anova.rare.MusaSpecies$`Pr(>F)`[1]
#Write to file
write.csv(file="Table_1.csv", Table.1, row.names=FALSE)
## Supplementary Figure 2 ##
#Combine habitat dataframes
disp.df <- rbind(disp.nmds.Habitat.summary, disp.rare.Habitat.summary)
#Set order for plot
disp.df$hab <- factor(disp.df$hab, levels=c("Roadside", "Jungle edge", "Jungle buffer", "Ravines"))
#Make dataframe for plot labels
disp.lab <- rbind(cbind(as.data.frame(metadata.nmds %>% group_by(Habitat.summary) %>% dplyr::summarise(n=n())),
data="nmds"),
cbind(as.data.frame(metadata.rare %>% group_by(Habitat.summary) %>% dplyr::summarise(n=n())),
data="rare"))
#Plot boxplot
gg.disp <- ggplot(disp.df, aes(x=distances, y=hab)) +
facet_wrap(. ~ data, labeller=labeller(data=c(`nmds`="Common OTUs", `rare`="All OTUs (including rare)"))) +
geom_boxploth(aes(fill=hab)) +
geom_text(data=disp.lab,
aes(x=Inf, y=Habitat.summary, label=paste0("n=", n)),
vjust=-0.5,
hjust=-0.25,
size=2.5,) +
labs(x="Distance from centroid") +
scale_fill_manual(values=c("#6FD4FF", "#F9FF00", "#79EC14", "#fb9a99"),
name="Habitat") +
theme_minimal() +
theme(axis.title.x=element_text(margin=unit(c(2, 0, 0, 0), "mm"), size=10),
axis.title.y=element_blank(),
axis.text=element_text(size=8),
panel.spacing.x=unit(0.5, "in"),
legend.position="top",
legend.title=element_text(size=6),
legend.text=element_text(size=5),
plot.margin=unit(c(10, 35, 0, 10), "pt")) +
coord_cartesian(clip="off")
#Write to file - SUPPLEMENTARY FIGURE 2
tiff(file=paste0("Supplementary_Figure_2-", Sys.Date(), ".tiff"), height=7, width=12, units="cm", res=300, compression="lzw")
gg.disp
dev.off()
#Plot NMDS with TZ fit
gg.nmds1 <- ggplot() +
geom_contour(data=nmds.ordi.tz.gg,
aes(x=x, y=y, z=z),
colour="dimgrey",
binwidth=5,
show.legend=FALSE,
size=0.3) +
geom_label_contour(data=nmds.ordi.tz.gg,
aes(x=x, y=y, z=z),
colour="dimgrey",