/
DiffIsoAnalysis.Rmd
367 lines (236 loc) · 13.7 KB
/
DiffIsoAnalysis.Rmd
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
---
title: "Total v Nuclear"
author: "Briana Mittleman"
date: "5/2/2019"
output: html_document
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```
```{r}
library(workflowr)
library(tidyverse)
library(reshape2)
```
In this analysis I wil use leafcutter to call PAS with differential ussage between fractions.
##Prepare annotation
I first filter the annotated peak SAF file for peaks passing the 5% coverage in either fraction.
```{bash,eval=F}
python makeSAFbothfrac5perc.py
```
##Peak quantification
```{bash,eval=F}
mkdir bothFrac_FC
```
Run feature counts with these peaks with both fractions:
```{bash,eval=F}
sbatch bothFrac_FC.sh
```
Fix the header:
```{bash,eval=F}
python fixFChead_bothfrac.py ../data/bothFrac_FC/APApeaks.ALLChrom.Filtered.Named.GeneLocAnnoPARSED.5percCov.bothfrac.fc ../data/bothFrac_FC/APApeaks.ALLChrom.Filtered.Named.GeneLocAnnoPARSED.5percCov.bothfrac.fixed.fc
```
Remove location demoniaiton:
#Prepare leafcutter phenotype
```{bash,eval=F}
mkdir ../data/DiffIso
```
```{bash,eval=F}
python fc2leafphen.py
```
Fix pheno to remove location:
```{bash,eval=F}
python removeloc_pheno.py ../data/DiffIso/APApeaks.ALLChrom.Filtered.Named.GeneLocAnnoPARSED.5percCov.bothfrac.fixed.forLC.fc ../data/DiffIso/APApeaks.ALLChrom.Filtered.Named.GeneLocAnnoPARSED.5percCov.bothfrac.fixed.forLC_noloc.fc
```
```{bash,eval=F}
python subset_diffisopheno.py 1
python subset_diffisopheno.py 2
python subset_diffisopheno.py 3
python subset_diffisopheno.py 4
python subset_diffisopheno.py 5
python subset_diffisopheno.py 6
python subset_diffisopheno.py 7
python subset_diffisopheno.py 8
python subset_diffisopheno.py 9
python subset_diffisopheno.py 10
python subset_diffisopheno.py 11
python subset_diffisopheno.py 12
python subset_diffisopheno.py 13
python subset_diffisopheno.py 14
python subset_diffisopheno.py 15
python subset_diffisopheno.py 16
python subset_diffisopheno.py 18
python subset_diffisopheno.py 19
python subset_diffisopheno.py 20
python subset_diffisopheno.py 21
python subset_diffisopheno.py 22
```
Make the sample groups file:
```{bash,eval=F}
python LC_samplegroups.py
```
##Run leafcutter
The leafcutter environment is not in the three-prime-seq environment. Make sure leafcutter is installed and working.
```{bash,eval=F}
sbatch run_leafcutterDiffIso.sh
```
Concatinate results:
```{bash,eval=F}
awk '{if(NR>1)print}' ../data/DiffIso/TN_diff_isoform_chr*.txt_effect_sizes.txt > ../data/DiffIso/TN_diff_isoform_allChrom.txt_effect_sizes.txt
awk '{if(NR>1)print}' ../data/DiffIso/TN_diff_isoform_chr*.txt_cluster_significance.txt > ../data/DiffIso/TN_diff_isoform_AllChrom_cluster_significance.txt
```
##Evaluate results
###Significant clusters
```{r}
sig=read.table("../data/DiffIso/TN_diff_isoform_AllChrom_cluster_significance.txt",sep="\t" ,col.names = c('status','loglr','df','p','cluster','p.adjust'),stringsAsFactors = F) %>% filter(status=="Success")
sig$p.adjust=as.numeric(as.character(sig$p.adjust))
```
```{r}
qqplot(-log10(runif(nrow(sig))), -log10(sig$p.adjust),ylab="-log10 Total Adjusted Leafcutter pvalue", xlab="-log 10 Uniform expectation", main="Leafcutter differencial isoform analysis between fractions")
abline(0,1)
```
```{r}
tested_genes=nrow(sig)
tested_genes
```
```{r}
sig_genes=sig %>% filter(p.adjust<.05)
number_sig_genes=nrow(sig_genes)
number_sig_genes
sig_genesonly=sig_genes %>% separate(cluster,into=c("chrom", "geneName"), sep = ":") %>% dplyr::select(geneName)
write.table(sig_genesonly, file="../data/sigDiffGenes.txt", col.names = T, row.names = F, quote = F)
```
###Effect sizes
```{r}
effectsize=read.table("../data/DiffIso/TN_diff_isoform_AllChrom_effect_sizes.txt", stringsAsFactors = F, col.names=c('intron', 'logef' ,'Nuclear', 'Total','deltaPAU')) %>% filter(intron != "intron")
write.table(effectsize,file="../data/DiffIso/EffectSizes.txt", quote = F, col.names = T, row.names = F)
effectsize$deltaPAU=as.numeric(as.character(effectsize$deltaPAU))
effectsize$logef=as.numeric(as.character(effectsize$logef))
```
Plot delta PAU:
```{r}
plot(sort(effectsize$deltaPAU),main="Leafcutter delta PAU", ylab="Delta PAU", xlab="PAS Index")
```
Filter PAU > .2
```{r}
effectsize_deltaPAU= effectsize %>% filter(abs(deltaPAU) > .2)
nrow(effectsize_deltaPAU)
effectSize_highdiffGenes=effectsize_deltaPAU %>% separate(intron, into=c("chrom", "start", "end", "GeneName"), sep=":") %>% dplyr::select(GeneName) %>% unique()
write.table(effectSize_highdiffGenes, file="../data/highdiffsiggenes.txt", col.names = F, row.names = F, quote = F)
```
Genes in this set:
```{r}
effectsize_deltaPAU_Genes= effectsize_deltaPAU %>% separate(intron, into=c("chrom", "start", "end","gene"),sep=":") %>% group_by(gene) %>% summarise(nperGene=n())
nrow(effectsize_deltaPAU_Genes)
```
Filter >.2 in Nuclear
```{r}
effectsize_deltaPAU_nuclear= effectsize_deltaPAU %>% filter(deltaPAU < -0.2)
```
Filter >.2 in Total:
```{r}
effectsize_deltaPAU_total= effectsize_deltaPAU %>% filter(deltaPAU > 0.2)
```
##Location of high >PAU
###Total:
Pull in location information for each PAS:
```{r}
PAS=read.table("../data/peaks_5perc/APApeaks.ALLChrom.Filtered.Named.GeneLocAnnoPARSED.5percCov.bothfrac.SAF",stringsAsFactors = F,header = T) %>% separate(GeneID, into=c("num", "chr", "start", "end", "strand", "geneID"), sep=":") %>% separate(geneID, into=c("gene", "loc"),sep="_") %>% mutate(intron=paste("chr", Chr, ":", Start, ":", End, ":", gene,sep="")) %>% select(intron, loc)
```
```{r}
effectsize_deltaPAU_total_loc=effectsize_deltaPAU_total %>% inner_join(PAS, by="intron")
ggplot(effectsize_deltaPAU_total_loc,aes(x=loc)) + geom_histogram(stat="count") + labs(title="Location of Total peaks >.2 PAU")
```
###Nuclear:
```{r}
effectsize_deltaPAU_nuclear_loc=effectsize_deltaPAU_nuclear %>% inner_join(PAS, by="intron")
ggplot(effectsize_deltaPAU_nuclear_loc,aes(x=loc)) + geom_histogram(stat="count") + labs(title="Location of Nuclear peaks >.2 PAU")
```
I will want to look at proportions. I need to know how many peaks are in each location:
```{r}
PAS_loc =PAS%>% group_by(loc) %>% summarise(nloc=n())
```
```{r}
effectsize_deltaPAU_total_locProp=effectsize_deltaPAU_total_loc %>% group_by(loc) %>% summarise(nloctotal=n())
effectsize_deltaPAU_nuclear_locProp=effectsize_deltaPAU_nuclear_loc %>% group_by(loc) %>% summarise(nlocnuclear=n())
effectsize_deltaPAUProp_tot=effectsize_deltaPAU_total_locProp %>% inner_join(PAS_loc, by="loc") %>% mutate(Proportion_tot=nloctotal/nloc)
effectsize_deltaPAUProp_nuc=effectsize_deltaPAU_nuclear_locProp %>% inner_join(PAS_loc, by="loc") %>% mutate(Proportion_nuc=nlocnuclear/nloc)
```
```{r}
ggplot(effectsize_deltaPAUProp_tot, aes(x=loc, y=Proportion_tot)) + geom_bar(stat="identity") + labs(y="Proportion of all called PAS", title="Location of high Total used PAS")
```
```{r}
ggplot(effectsize_deltaPAUProp_nuc, aes(x=loc, y=Proportion_nuc)) + geom_bar(stat="identity") + labs(y="Proportion of all called PAS", title="Location of high nuclear used PAS")
```
Merge to 1 figure:
```{r}
effectsize_deltaPAUProp_both= effectsize_deltaPAUProp_nuc %>% inner_join(effectsize_deltaPAUProp_tot, by=c("loc","nloc")) %>% rename(Nuclear=Proportion_nuc, Total=Proportion_tot) %>% select(loc, Nuclear, Total)
effectsize_deltaPAUProp_both_melt= effectsize_deltaPAUProp_both %>% melt(id.vars="loc", variable.name="Fraction", value.name = "Proportion")
```
```{r}
ggplot(effectsize_deltaPAUProp_both_melt, aes(x=loc, y=Proportion, by=Fraction, fill=Fraction)) + geom_bar(stat="identity", position="dodge") + scale_fill_manual(values=c("deepskyblue3","darkviolet")) + labs(title="Proportion of PAS differentiall used by location")
```
More differentiall used in total. this makes sense because there are more used peaks in the nuclear which evens out the distribution of the ratios.
###Stratify by different $\Delta$ PAU
I want to create a data frame that has the location proportion distribution based on different $\Delta$ PAU. 0-.1 .1-.2 .2-.3 .3-.4 .4-.5 >.5
First I will seperate the total and nuclear but the sign of the $\Delta$ PAU.
```{r}
Total_dpau= effectsize %>% filter(deltaPAU > 0) %>% inner_join(PAS, by="intron") %>% select(-logef, -Nuclear,-Total) %>% mutate(fraction="Total", PAU_Cat=ifelse(deltaPAU <.1, "<.1", ifelse(deltaPAU >=.1 & deltaPAU <.2, "<.2", ifelse(deltaPAU >=.2 & deltaPAU <.3, "<.3", ifelse(deltaPAU >=.3 & deltaPAU <.4, "<.4", "<.5")))))
Nuclear_dpau= effectsize %>% filter(deltaPAU <0) %>% inner_join(PAS, by="intron") %>% select(-logef,-Nuclear,-Total) %>% mutate(fraction="Nuclear", PAU_Cat=ifelse(deltaPAU >-.1, "<.1", ifelse(deltaPAU <=-.1 & deltaPAU > -.2, "<.2", ifelse(deltaPAU <=-.2 & deltaPAU >-.3, "<.3", ifelse(deltaPAU <=-.3 & deltaPAU >-.4, "<.4", "<.5")))))
```
Merge these together to start grouping:
```{r}
allPAU=as.data.frame(rbind(Total_dpau, Nuclear_dpau)) %>% group_by(fraction, PAU_Cat, loc ) %>% summarise(nperLoc=n()) %>% full_join(PAS_loc, by ="loc") %>% mutate(Prop=nperLoc/nloc)
```
Plot it:
```{r}
ggplot(allPAU, aes(x=loc,y=Prop, group=fraction, fill=fraction)) + geom_bar(stat="identity", position = "dodge") + facet_wrap(~PAU_Cat)+ scale_fill_manual(values=c("deepskyblue3","darkviolet")) + theme(axis.text.x = element_text(angle = 90, hjust = 1)) + labs(title="Proportion of PAS by location and delta PAU")
```
```{r}
allPAU_remove.1= allPAU %>% filter(PAU_Cat != "<.1")
ggplot(allPAU_remove.1, aes(x=loc,y=Prop, group=fraction, fill=fraction)) + geom_bar(stat="identity", position = "dodge") + facet_wrap(~PAU_Cat)+ scale_fill_manual(values=c("deepskyblue3","darkviolet")) + theme(axis.text.x = element_text(angle = 90, hjust = 1)) + labs(title="Proportion of PAS by location and delta PAU")
```
Proportion within group:
```{r}
allPAU_ingroup= allPAU %>% mutate(nCat=sum(nperLoc),proppercat=nperLoc/nCat)
ggplot(allPAU_ingroup, aes(x=loc,y=proppercat, group=fraction, fill=fraction)) + geom_bar(stat="identity", position = "dodge") + facet_wrap(~PAU_Cat)+ scale_fill_manual(values=c("deepskyblue3","darkviolet")) + theme(axis.text.x = element_text(angle = 90, hjust = 1)) + labs(title="Proportion of PAS by location and delta PAU")
```
##Distance to TSS:
I need to pull in the TSS information so I can look at the distance between the differentially used peaks and by distance .
```{r}
tss=read.table("../../genome_anotation_data/refseq.ProteinCoding.bed",col.names = c("chrom", "start", "end", "gene", "score", "strand") ,stringsAsFactors = F) %>% mutate(TSS= ifelse(strand=="+", start, end)) %>% select(gene, TSS, strand)
```
Seperate effect size introns:
PAS base for + strand is end, PAS for neg stand in -
```{r}
effectsize_TSS= effectsize %>% separate(intron, into=c("chrom", "start", "end", "gene"),sep=":") %>% mutate(fraction=ifelse(deltaPAU < 0, "nuclear", "total")) %>% inner_join(tss, by="gene") %>% mutate(dist2PAS=ifelse(strand=="+", as.numeric(end)-as.numeric(TSS), as.numeric(TSS)-as.numeric(start)))
effectsize_TSS_tot= effectsize_TSS %>% filter(fraction=="total") %>% mutate( PAU_Cat=ifelse(deltaPAU <.1, "<.1", ifelse(deltaPAU >=.1 & deltaPAU <.2, "<.2", ifelse(deltaPAU >=.2 & deltaPAU <.3, "<.3", ifelse(deltaPAU >=.3 & deltaPAU <.4, "<.4", "<.5")))))
effectsize_TSS_nuc=effectsize_TSS %>% filter(fraction=="nuclear") %>% mutate( PAU_Cat=ifelse(deltaPAU >-.1, "<.1", ifelse(deltaPAU <=-.1 & deltaPAU > -.2, "<.2", ifelse(deltaPAU <=-.2 & deltaPAU >-.3, "<.3", ifelse(deltaPAU <=-.3 & deltaPAU >-.4, "<.4", "<.5")))))
effectsize_TSS_cat=as.data.frame(rbind(effectsize_TSS_tot, effectsize_TSS_nuc)) %>% filter(dist2PAS >0)
```
```{r}
ggplot(effectsize_TSS_cat, aes(x=log10(dist2PAS), by=fraction, fill=fraction))+ geom_density(alpha=.4) + facet_grid(~PAU_Cat) + labs(title="Distance to TSS for differentialy used PAS")+scale_fill_manual(values=c("deepskyblue3","darkviolet"))
```
###By length of gene
```{r}
length=read.table("../../genome_anotation_data/refseq.ProteinCoding.bed",col.names = c("chrom", "start", "end", "gene", "score", "strand") ,stringsAsFactors = F) %>% mutate(length=abs(end-start)) %>% mutate(TSS= ifelse(strand=="+", start, end)) %>% select(gene, length,TSS, strand)
```
```{r}
effectsize_length= effectsize %>% separate(intron, into=c("chrom", "start", "end", "gene"),sep=":") %>% mutate(fraction=ifelse(deltaPAU < 0, "nuclear", "total")) %>% inner_join(length, by="gene") %>% mutate(PercLength=ifelse(strand=="+", ((as.numeric(end)-as.numeric(TSS))/as.numeric(length)), (1-(as.numeric(start)-as.numeric(TSS))/as.numeric(length))))
effectsize_length_tot= effectsize_length %>% filter(fraction=="total") %>% mutate( PAU_Cat=ifelse(deltaPAU <.1, "<.1", ifelse(deltaPAU >=.1 & deltaPAU <.2, "<.2", ifelse(deltaPAU >=.2 & deltaPAU <.3, "<.3", ifelse(deltaPAU >=.3 & deltaPAU <.4, "<.4", "<.5")))))
effectsize_length_nuc=effectsize_length %>% filter(fraction=="nuclear") %>% mutate( PAU_Cat=ifelse(deltaPAU >-.1, "<.1", ifelse(deltaPAU <=-.1 & deltaPAU > -.2, "<.2", ifelse(deltaPAU <=-.2 & deltaPAU >-.3, "<.3", ifelse(deltaPAU <=-.3 & deltaPAU >-.4, "<.4", "<.5")))))
effectsize_length_cat=as.data.frame(rbind(effectsize_length_tot, effectsize_length_nuc)) %>% filter(PercLength<=1 & PercLength >0)
effectsize_length_catall=as.data.frame(rbind(effectsize_length_tot, effectsize_length_nuc))
```
```{r}
ggplot(effectsize_length_cat, aes(x=PercLength, by=fraction, fill=fraction))+ geom_histogram(alpha=.4,bins=10) + facet_grid(~PAU_Cat) + labs(title="Location of differentially used PAS within a gene body ")+scale_fill_manual(values=c("deepskyblue3","darkviolet"))
```
```{r}
summary(effectsize_length_catall$PercLength)
```
```{r}
summary(effectsize$logef)
```
```{r}
ggplot(effectsize_length_cat, aes(x=PercLength, by=fraction, fill=fraction))+ geom_histogram(alpha=.4,bins=50) + labs(title="Location of differentially used PAS \nwithin a gene body ")+scale_fill_manual(values=c("deepskyblue3","darkviolet"))
```