Missing Transcriptomics raw data

3 of 90 transcriptomics samples are missing on SRA.

The remaining 87 were downloaded, zipped and uploaded to the respective ARC assay using:

• command line tool: sratools
• dir: sratoolkit.3.0.0-win64/bin
• command: .\prefetch SRR1548xxxx ; .\fasterq-dump -O .\SRR1548xxxx SRR1548xxxx -p

The missing files were reported to Erin and the SRA office:

• recovery_24h_35C_C (SRX11781333)
• heat_0.5h_40C_A (SRX11781343)
• heat_16h_40C_C (SRX11781360)

changed the description

Erin provided me with the three missing fastq.gz files. These are unprocessed paired end files merged in a single file. I couldn't find a tool to de-interlace these files into individual fastq files. Galaxy's fastq de-interlacer only works for old versions of fastq (/1 rather than the new 1:). I wrote a simple script for the de-interlacing. It is absolutely not efficient, but works.

As soon as the files are available at SRA, we should replace the files with the SRA-versions!

Simple script for

#r "nuget: FSharpAux.IO, 1.1.0"

open FSharpAux.IO

let path = 
    @"\\syno.bio.uni-kl.de\CSB\RuZhang\Transcripts\sratoolkit.3.0.0-win64\bin\35C_C_14.fastq"
    //@"\\syno.bio.uni-kl.de\CSB\RuZhang\Transcripts\sratoolkit.3.0.0-win64\bin\40C_A_3.fastq"

let read = FSharpAux.IO.FileIO.readFile path

let mutable collR1 : string list = []
let mutable collR2 : string list = []
let mutable counterR1 = 0
let mutable counterR2 = 0

let mutable index = 1

let e = read.GetEnumerator()

let newR1 = 
    @"\\syno.bio.uni-kl.de\CSB\RuZhang\Transcripts\sratoolkit.3.0.0-win64\bin\35C_C_14_R1.fastq"
    //@"\\syno.bio.uni-kl.de\CSB\RuZhang\Transcripts\sratoolkit.3.0.0-win64\bin\40C_A_3_R1.fastq"

let newR2 = 
    @"\\syno.bio.uni-kl.de\CSB\RuZhang\Transcripts\sratoolkit.3.0.0-win64\bin\35C_C_14_R2.fastq"
    //@"\\syno.bio.uni-kl.de\CSB\RuZhang\Transcripts\sratoolkit.3.0.0-win64\bin\40C_A_3_R2.fastq"

while e.MoveNext() do
    let tmp :string = e.Current
    if tmp.StartsWith "@" then
        if counterR2%1_000_000=0 then
            if  counterR1%1_000_000=0 then 
                FileIO.writeToFile true newR1 (collR1 |> List.rev)
                collR1 <- []
                FileIO.writeToFile true newR2 (collR2 |> List.rev)
                collR2 <- []
                printfn "R1 %i\nR2 %i" counterR1 counterR2

        if tmp.Contains " 1:" then
            index <- 1
            counterR1 <- counterR1 + 1
            collR1 <- tmp::collR1

        elif tmp.Contains " 2:"then
            index <- 2
            counterR2 <- counterR2 + 1
            collR2 <- tmp::collR2

        else failwithf "no :1/:2 in  %s" tmp
    else 
        if index = 1 then
            collR1 <- tmp::collR1
        else collR2 <- tmp::collR2

//write the remaining chunk to the file
FileIO.writeToFile true newR1 (collR1 |> List.rev)
FileIO.writeToFile true newR2 (collR2 |> List.rev)


printfn "index should be 2 and it is: %i" index       //should be 2
printfn "Counter should be equal\nR1Counter: %i\nR2Counter: %i" counterR1 counterR2

The temporary files are located here:

• recovery_24h_35C_C (SRX11781333) -> "\assays\Transcriptomics\dataset\fastq\35C\recovery_24h\35C_C"
• heat_0.5h_40C_A (SRX11781343) -> "\assays\Transcriptomics\dataset\fastq\40C\heat_00.5h\40C_A"
• heat_16h_40C_C (SRX11781360) -> "assays\Transcriptomics\dataset\fastq\40C\heat_16h\40C_C"

35850338

The missing samples are now available at SRA, but the format differs from the remaining samples. The files are not deinterlaced using fasterq-dump, but remain as single fastq file. Since I already got the data and deinterlaced them manually, I close this issue.

sample	command
recovery_24h_35C_C:	.\prefetch SRR15481355 ; .\fasterq-dump -O .\SRR15481355 SRR15481355 -p
heat_0.5h_40C_A	.\prefetch SRR15481365 ; .\fasterq-dump -O .\SRR15481365 SRR15481365 -p
heat_16h_40C_C	.\prefetch SRR15481382 ; .\fasterq-dump -O .\SRR15481382 SRR15481382 -p

closed

mentioned in commit 1aadf2ad