makeCNVcalls.R failed

[lmanchon]

failed again with:

[1] "BEGIN makeCNVCalls.R"
[1] "Processing sample: SRR6823378_recalibrated 1/5"
Optimization of the choice of aggregate reference set, this process can take some time
Number of selected bins: 10000
Now fitting the beta-binomial model on a data frame with 206472 rows : this step can take a few minutes.
Now computing the likelihood for the different copy number states
Error in CallCNVs(x = all.exons, transition.probability = trans_prob, :
Chromosome, start and end vector must have the same lengths.
Calls: CallCNVs -> CallCNVs
In addition: Warning messages:
1: In aod::betabin(data = data.for.fit, formula = as.formula(formula), :
The data set contains at least one line with weight = 0.

2: In aod::betabin(data = data.for.fit, formula = as.formula(formula), :
The data set contains at least one line with weight = 0.

3: In aod::betabin(data = data.for.fit, formula = as.formula(formula), :
The data set contains at least one line with weight = 0.

Execution halted

[Kiran936]

Can you share your command and head of the input CustoExon file and First Step BED file..

Note: I am not a developer of this.. i could run without error,, i had errors bcz of input CustExon file

[lmanchon]

Rscript ./DECoN/Linux/ReadInBams.R --bams ./bamfiles --bed exome.bed --fasta Homo_sapiens_hg38.fasta
Rscript ./DECoN/Linux/IdentifyFailures.R --Rdata DECoNBams.RData --mincorr .98 --mincov 100 --custom FALSE
Rscript ./DECoN/Linux/makeCNVcalls.R --transProb 0.01 --Rdata DECoNBams.RData --plot All --exon exome.bed --custom TRUE --plotFolder out_F

more exome.bed
chr1 69090 70008
chr1 450739 451678
chr1 685715 686654
chr1 925941 926013
chr1 930154 930336
chr1 931038 931089
chr1 935771 935896
chr1 939039 939129
chr1 939274 939460
chr1 941143 941306
chr1 942135 942251
chr1 942409 942488
chr1 942558 943058
chr1 943252 943377
chr1 943697 943808
chr1 943907 944153
chr1 944693 944800
chr1 945056 945146
chr1 945517 945653
chr1 946172 946286
chr1 946401 946545
chr1 948130 948232

[Kiran936]

I see you are giving Custom True and your custom bed file do not have exon numbering on 4th Column with headers in makeCNVcalls step,

1. Intersect your exome bed with genes. in First step and pass it in ReadInBam, ( Chr Start End Gene )
2. Make another exome/Custom number bed file Chr Start End Gene Exon.Number this must be passed in MakeCNVCalls step

[lmanchon]

this is my new bed file below and my .Rdata file in attachment

chr Start end Gene
chr1 11868 12227 DDX11L1
chr1 12612 12721 DDX11L1
chr1 13220 14409 DDX11L1
chr1 14403 14501 WASH7P
chr1 15004 15038 WASH7P
chr1 15795 15947 WASH7P
chr1 16606 16765 WASH7P
chr1 16857 17055 WASH7P
chr1 17232 17368 WASH7P
chr1 17368 17436 MIR6859-1
chr1 17605 17742 WASH7P
chr1 17914 18061 WASH7P
chr1 18267 18366 WASH7P
chr1 24737 24891 WASH7P
chr1 29533 29570 WASH7P
chr1 29553 30039 MIR1302-2HG
chr1 30365 30503 MIR1302-2
chr1 30563 30667 MIR1302-2HG
chr1 30975 31097 MIR1302-2HG
chr1 34553 35174 FAM138A
chr1 35276 35481 FAM138A
chr1 35720 36081 FAM138A
chr1 52472 53312 OR4G4P
chr1 57597 57653 OR4G11P
chr1 58699 58856 OR4G11P
chr1 62915 64116 OR4G11P
chr1 69054 70108 OR4F5
chr1 89294 91629 AL627309.1
chr1 89550 90050 AL627309.3
chr1 90286 91105 AL627309.3
DECoNBams.zip

[Kiran936]

and create custom exon number file with same BED file as above.. Chr Start End Gene Cust.Exonumber then pass it to Make CNV Step

[lmanchon]

Exon.Number means the number of exons per gene, isn't it ?

[Kiran936]

Chr Start End Gene Cust.ExonNo
chr1 11873 12227 DDX11L1 1
chr1 12594 12721 DDX11L1 2
chr1 13402 14409 DDX11L1 3
chr1 14406 16765 WASH7P 1
chr1 14969 15038 WASH7P 2
chr1 15795 15947 WASH7P 3
chr1 16606 16765 WASH7P 4

Its exon numbering for each exon of gene.. if 18 Exons per gene,, then 1-18 exons numbers

[lmanchon]

ok, it is the position number of exon in gene

[lmanchon]

ok, it failed at first loop step in makeCNVcalls.R with:

Optimization of the choice of aggregate reference set, this process can take some time
Number of selected bins: 10000
Now fitting the beta-binomial model on a data frame with 198253 rows : this step can take a few minutes.
Now computing the likelihood for the different copy number states
Error in CallCNVs(x = all.exons, transition.probability = trans_prob, :
Chromosome, start and end vector must have the same lengths.
In addition: Warning messages:
1: In aod::betabin(data = data.for.fit, formula = as.formula(formula), :
The data set contains at least one line with weight = 0.

2: In aod::betabin(data = data.for.fit, formula = as.formula(formula), :
The data set contains at least one line with weight = 0.

3: In aod::betabin(data = data.for.fit, formula = as.formula(formula), :
The data set contains at least one line with weight = 0.

4: In aod::betabin(data = data.for.fit, formula = as.formula(formula), :
The data set contains at least one line with weight = 0.

[Kiran936]

Something is wrong with your CustomExon Bed file, either format or the numbering, also check chr naming from your BAM, and both of BED files from step1 and Custom Exon

[Kiran936]

you can omit the custom exon file from MakeCNV,, thats an optional try without that