I have big peak list in the "Bed" format and I converted it to GenomicRange for use as an input for the GADEM package to find denovo motifs. But when I try the GADEM function always I face the below error.
Could you please anybody who knows help me with this error?
This is a small example of my real file with only 20 rows.
1 chr6 29723590 29723790
2 chr14 103334312 103334512
3 chr1 150579030 150579230
4 chr7 76358527 76358727
5 chr6 11537891 11538091
6 chr14 49893256 49893456
7 chr5 179623200 179623400
8 chr1 228082831 228083031
9 chr12 93441644 93441844
10 chr10 3784776 3784976
11 chr3 183635833 183636033
12 chr7 975301 975501
13 chr12 123364510 123364710
14 chr1 1615578 1615778
15 chr1 36156320 36156520
16 chr14 55051781 55051981
17 chr8 11867697 11867897
18 chr22 38706135 38706335
19 chr6 44265256 44265456
20 chr1 185316658 185316858
and the code that I use is :
library(GenomicRanges)
library(rGADEM)
data = makeGRangesFromDataFrame(data, keep.extra.columns = TRUE)
data = reduce(data)
data = resize(data, width = 50, fix='center')
gadem<-GADEM(data,verbose=1,genome=Hsapiens)
plot(gadem)
and error is:
[ Retrieving sequences... Error in.Call2("C_solve_user_SEW", refwidths, start, end, width, translate.negative.coord:
solving row 136: 'allow.nonnarrowing' is FALSE and the supplied start (55134751) is > refwidth + 1 ]
Better to mention that, when I try an example input file with less than 136 rows, it works and I get motifs.
Thanks in advance.
Related
I am trying to use the RCircos package in R to visualize links between genomic positions. I am unfamiliar with this package and have been using the package documentation available from the CRAN repository from 2016.
I have attempted to format my data according to the package requirements. Here is what it looks like:
> head(pts3)
Chromosome ChromStart ChromEnd Chromosome.1 ChromStart.1 ChromEnd.1
1 chr1 33 34 chr1 216 217
2 chr1 33 34 chr1 789 790
3 chr1 33 34 chr1 1716 1717
4 chr1 33 34 chr1 1902 1903
5 chr1 33 34 chr2 2538 2539
6 chr1 33 34 chr2 4278 4279
Ultimately, I would like to produce a plot with tracks from ChromStart to ChromStart.1 and each gene labeled along the outside of the plot. I thought the script would look something like:
RCircos.Set.Core.Components(cyto.info = pts3,
chr.exclude = NULL,
tracks.inside = 1,
tracks.outside = 2)
RCircos.Set.Plot.Area()
RCircos.Chromosome.Ideogram.Plot()
RCircos.Link.Plot(link.data = pts3,
track.num = 3,
by.chromosome = FALSE)
It appears that to do so, I must first initialize with the RCircos.Set.Core.Components() function which requires positional information for each gene to pass to RCircos.Chromosome.Ideogram.Plot(). So, I created a second data frame containing the required information to pass to the function and this is the error that I get:
> head(genes)
Chromosome ChromStart ChromEnd GeneName Band Stain
1 chr1 0 2342 PB2 NA NA
2 chr2 2343 4683 PB1 NA NA
3 chr3 4684 6917 PA NA NA
4 chr4 6918 8710 HA NA NA
5 chr5 8711 10276 NP NA NA
6 chr6 10277 11735 NA NA NA
> RCircos.Set.Core.Components(cyto.info = genes,
+ chr.exclude = NULL,
+ tracks.inside = 1,
+ tracks.outside = 2)
Error in RCircos.Validate.Cyto.Info(cyto.info, chr.exclude) :
Cytoband start should be 0.
I don't actually have data for the Band or Stain columns and don't understand what they are for, but adding data to the those columns (such as 1:8 or chr1, chr2, etc) does not resolve the problem. Based on a recommendation from another forum, I also tried to reset the plot parameters for RCircos using the following functions, but it did not resolve the error:
core.chrom <- data.frame("Chromosome" = c("chr1", "chr2", "chr3", "chr4", "chr5", "chr6", "chr7", "chr8"),
"ChromStart" = c(0, 2343, 4684, 6918, 8711, 10277, 11736, 12763),
"ChromEnd" = c(2342, 4683, 6917, 8710, 10276, 11735, 12762, 13666),
"startLoc" = c(0, 2343, 4684, 6918, 8711, 10277, 11736, 12763),
"endLoc" = c(2342, 4683, 6917, 8710, 10276, 11735, 12762, 13666),
"Band" = NA,
"Stain" = NA)
RCircos.Reset.Plot.Ideogram(chrom.ideo = core.chrom)
Any advice would be deeply appreciated!
I'm not sure if you figured this one out or moved on etc. I had the same problem and ended up resolving it by reformatting my start positions for each chromosome to 0 as opposed to a continuation of the previous chr. For you it would be:
Chromosome ChromStart ChromEnd GeneName Band Stain
1 chr1 0 2342 PB2 NA NA
2 chr2 0 2340 PB1 NA NA
3 chr3 0 2233 PA NA NA
...etc
I have the following biological data file.
#acgh_file
chromosome startPosition
chr1 37196
chr1 52308
chr1 357503
chr1 443361
chr1 530358
and I need to convert the positions by means of a translation table.
#convert
chr1 37196 chr1 47333
chr1 52308 chr1 62445
chr1 357503 chr1 367640
chr1 443361 chr1 453498
chr1 530358 chr1 540495
What needs to happen is that I have to replace the startPosition in the acgh_file with the value in fourth column of the convert table.
I made a script, but as the files are quite large it takes ages before it finishes (probably due to that R is not good for doing for-loops).
for (n in 1:nrow(convert)){
acgh_file[acgh_file$chromosome==convert[n,1] & acgh_file$startPosition==convert[n,2],3] <- convert[n,4]
}
I'm looking for a quicker solution here. Anybody have some ideas? I thought about doing something with the apply functions, but I don't know how to combine that when using this convert look-up table that I have here.
No need to use a for-loop here( Btw for loop in R are slow when they are not used in the good manner). Here you want to do a merge between 2 data sets. Since you have a big data.frame, I suggest to use data.table package to do the merge.
library(data.table)
setkey(acgh_file,chromosome,startPosition)
setkey(convert_file,V1,V2)
acgh_file[convert_file]
# chromosome startPosition V4
# 1: chr1 37196 47333
# 2: chr1 52308 62445
# 3: chr1 357503 367640
# 4: chr1 443361 453498
# 5: chr1 530358 540495
where the data sets are data.table :
acgh_file <- fread("
chromosome startPosition
chr1 37196
chr1 52308
chr1 357503
chr1 443361
chr1 530358")
convert_file <- fread("
chr1 37196 chr1 47333
chr1 52308 chr1 62445
chr1 357503 chr1 367640
chr1 443361 chr1 453498
chr1 530358 chr1 540495")[,V3:=NULL]
I have a data frame with over 1000 rows.
1 chr1 23373262 23390014 NM_001081079 - Ogfrl1
10 chr1 43807075 43884553 NM_026430 - Uxs1
100 chr10 77069342 77081076 NM_001301671 + Sumo3
1000 chr5 137736931 137748902 NM_031405 - Srrt
1001 chr5 137737851 137737916 NR_106003 -
1002 chr5 137751126 137755469 NM_011639 - Trip6
1003 chr5 137755785 137774810 NM_031406 - Slc12a9
1004 chr5 138220145 138221335 NM_027242 + Ppp1r35
1005 chr5 138223133 138227929 NM_144913 - Mepce
1006 chr5 138229029 138263849 NM_001005426 + Zcwpw1
I want to sort them by row.names, which the the number in the first column above. My command is
mef_genes<-mef_genes[sort(as.integer(row.names(mef_genes)),decreasing=F),]
Its ranked as shown above. I want to rank like normally ascending, i.e 1,2,3...
Can anyone tell me how to do it?
Thank you very much
Try order instead of sort:
mef_genes <- mef_genes[order(as.integer(row.names(mef_genes))),]
I have a huge dataframe df which includes information about overlapping intervals (A) and (B) and on which chromosome (chrom) they were located. There is also information about a value (level of gene expression) observed over interval (A).
chrom value Astart Aend Bstart Bend
chr1 0 0 54519752 17408 17431
chr1 0 0 54519752 17368 17391
chr1 0 0 54519752 567761 567783
chr11 0 2 93466832 568111 568133
chr11 0 2 93466832 568149 568171
chr11 0 2 93466832 1880734 1880756
chr11 4 93466844 93466880 93466856 93466878
chr11 2 93466885 135006516 93466889 93466911
chr11 2 93466885 135006516 94199710 94199732
Note that the same interval may appear several times, for instance, an interval (B) will have been reported two times if it overlapped with two (A) intervals:
Astart(1)=========================Aend1 Astart(2)========================Aend(2)
Bstart(1)=======================================Bend(1)
chrom value Astart Aend Bstart Bend
chr1 0 0 25 15 35 #A(1) and B(1) overlap
chr1 1 28 45 15 35 #A(2) and B(1) overlap
Likewise, an interval (A) will have been reported two or more times if it overlapped with two or more (B) intervals:
Astart(3)===================================================================Aend(3)
Bstart(2)=========Bend(2) Bstart(3)===========Bend(3) Bstart(4)===============Bend(4)
chrom value Astart Aend Bstart Bend
chr4 0 10 100 15 25 #A(3) and B(2) overlap
chr4 0 10 100 30 75 #A(3) and B(3) overlap
chr4 3 10 100 80 120 #A(3) and B(4) overlap
My goal is to output all the individual positions from intervals (B) and the corresponding values from (A). I have a piece of code that beautifully outputs all the relevant positions in (B):
position <- unlist(mapply(seq, ans$Bstart, ans$Bend - 1))
> head(position)
[1] 17408 17409 17410 17411 17412 17413
The problem with this is that it is not enough to retrieve the chromosome information back from there. I need to check chromosome information AND position at the same time when I list these positions. That is because the same position integer may occur on several chromosomes, so I can't afterwards just run something like for position %in% range(Astart, Aend) output $chrom, $value (dummy code).
How can I retrieve (chrom, position, value) at the same time?
The expected result would be something like this:
> head(expected_result)
chrom position value
chr1 17408 0
chr1 17409 0
chr1 17410 0
chr1 17411 0
chr1 17412 0
chr1 17413 0
#skipping some lines to show another part of the dataframe
chr11 93466856 4
chr11 93466857 4
A call to ddply might be more elegant, but the logic would be the same:
dfA = read.table(textConnection("chrom value Astart Aend Bstart Bend
chr1 0 0 54519752 17408 17431
chr1 0 0 54519752 17368 17391
chr1 0 0 54519752 567761 567783
chr11 0 2 93466832 568111 568133
chr11 0 2 93466832 568149 568171
chr11 0 2 93466832 1880734 1880756
chr11 4 93466844 93466880 93466856 93466878
chr11 2 93466885 135006516 93466889 93466911
chr11 2 93466885 135006516 94199710 94199732"), header = TRUE)
dfB = as.data.frame(do.call(rbind,
apply(dfA, MARGIN = 1, FUN = function(x) {
cbind(mapply(seq,
as.numeric(x['Bstart']),
as.numeric(x['Bend']) - 1),
x['chrom'], x['value'])
}
)))
lapply(dfB, typeof)
I would like to apply a function that returns a matrix to each row of a large data.table object (original file is around 30 GB, I have 80 GB ram), and get back a data.table object. I'd like to do it efficiently. My current approach is the following:
my.function <- function(x){
alnRanges<-cigarToIRanges(x[6]);
alnStarts<-start(alnRanges)+as.numeric(x[4])-1;
alnEnds<-end(alnRanges)+as.numeric(x[4])-1;
y<-x[-4];
ys<-matrix(rep(y,length(alnRanges)),nrow=length(alnRanges),ncol=length(y),byrow=TRUE);
ys<-cbind(ys,alnStarts,alnEnds);
return(ys); # ys is a matrix
}
my.dt<-fread(my.file.name);
my.list.of.matrices<-apply(my.dt,1,my.function);
new.df<-do.call(rbind.data.frame,my.list.of.matrices);
colnames(new.df)[1:14]<-colnames(my.dt)[-4];
new.dt<-as.data.table(new.df);
Note1: I specify the my.function just to show that it returns a matrix, and that my apply line is therefore a list of matrices.
Note2: I am not sure how slow are the operations I am doing but seems that I could reduce the number of lines. For example, is it slow to convert a data frame to a data table for large objects?
Reproducible example:
Note that Arun and Roland made me think harder about the problem so I am still working on it... may be that I do not need these lines...
I want to take a sam file, and then create a new coordinates file where each read is split according to its CIGAR field.
My sam file:
qname rname pos cigar
2218 chr1 24613476 42M2S
2067 chr1 87221030 44M
2129 chr1 79702717 44M
2165 chr1 43113438 44M
2086 chr1 52155089 4M921N40M
code:
library("data.table");
library("GenomicRanges");
sam2bed <- function(x){
alnRanges<-cigarToIRanges(x[4]);
alnStarts<-start(alnRanges)+as.numeric(x[3])-1;
alnEnds<-end(alnRanges)+as.numeric(x[3])-1;
#y<-as.data.frame(x[,pos:=NULL]);
#ys<-y[rep(seq_len(nrow(y)),length(alnRanges)),];
y<-x[-3];
ys<-matrix(rep(y,length(alnRanges)),nrow=length(alnRanges),ncol=length(y),byrow=TRUE);
ys<-cbind(ys,alnStarts,alnEnds);
return(ys);
}
sam.chr.dt<-fread(sam.parent.chr.file);
setnames(sam.chr.dt,old=c("V1","V2","V3","V4"),new=c("qname","rname","pos","cigar"));
bed.chr.lom<-apply(sam.chr.dt,1,sam2bed);
> bed.chr.lom
[[1]]
alnStarts alnEnds
[1,] "2218" "chr1" "42M2S" "24613476" "24613517"
[[2]]
alnStarts alnEnds
[1,] "2067" "chr1" "44M" "87221030" "87221073"
[[3]]
alnStarts alnEnds
[1,] "2129" "chr1" "44M" "79702717" "79702760"
[[4]]
alnStarts alnEnds
[1,] "2165" "chr1" "44M" "43113438" "43113481"
[[5]]
alnStarts alnEnds
[1,] "2086" "chr1" "4M921N40M" "52155089" "52155092"
[2,] "2086" "chr1" "4M921N40M" "52156014" "52156053"
bed.chr.df<-do.call(rbind.data.frame,bed.chr.lom);
> bed.chr.df
V1 V2 V3 alnStarts alnEnds
1 2218 chr1 42M2S 24613476 24613517
2 2067 chr1 44M 87221030 87221073
3 2129 chr1 44M 79702717 79702760
4 2165 chr1 44M 43113438 43113481
5 2086 chr1 4M921N40M 52155089 52155092
6 2086 chr1 4M921N40M 52156014 52156053
bed.chr.dt<-as.data.table(bed.chr.df);
> bed.chr.dt
V1 V2 V3 alnStarts alnEnds
1: 2218 chr1 42M2S 24613476 24613517
2: 2067 chr1 44M 87221030 87221073
3: 2129 chr1 44M 79702717 79702760
4: 2165 chr1 44M 43113438 43113481
5: 2086 chr1 4M921N40M 52155089 52155092
6: 2086 chr1 4M921N40M 52156014 52156053
Assuming ff is your data.table, how about this?
splits <- cigarToIRangesListByAlignment(ff$cigar, ff$pos, reduce.ranges = TRUE)
widths <- width(attr(splits, 'partitioning'))
cbind(data.table(qname=rep.int(ff$qname, widths),
rname=rep.int(ff$rname, widths)), as.data.frame(splits))
qname rname space start end width
1: 2218 chr1 1 24613476 24613517 42
2: 2067 chr1 2 87221030 87221073 44
3: 2129 chr1 3 79702717 79702760 44
4: 2165 chr1 4 43113438 43113481 44
5: 2086 chr1 5 52155089 52155092 4
6: 2086 chr1 5 52156014 52156053 40