Creating tx2gene table from sheep transcripts IDs to gene IDs using NCBI annotation file

Question

[Sacit]

Hello,

I am trying to create a tx2gene table using NCBI annotation file (GCF_016772045.1) - used following codes:

ramb_gtf <- import("genomic.gtf")
txdb <- makeTxDbFromGRanges(ramb_gtf)
k <- keys(txdb, keytype = "GENEID")
df <- select(txdb, keys = k,  columns = "TXNAME", keytype = "GENEID")
tx2gene <- df[, 2:1]
head(tx2gene)

I got following tx2gene header:

TXNAME GENEID
1 XM_027978951.2   A1BG
2 XM_004020007.5   A1CF
3 XM_004020008.5   A1CF
4 XM_004006891.4  A2ML1

My salmon quant files looks like:

Name    Length  EffectiveLength TPM     NumReads
NM_001009196.1  409     205.307 0.000000        0.000
NM_001009196.1  409     205.307 0.000000        0.000
NM_001009201.1  1110    905.814 0.000000        0.000
NM_001009202.1  1102    897.814 5428.512703     3.000
NM_001009204.1  201     60.517  0.000000        0.000

Am I using different annotation file or there is some thing different?

PS: I am newbie

[Pierre Lindenbaum]

You have the code, you have a salmon output. What is your problem ?

[Sacit]

tx2gene table have XM_ (predicted) IDs while quant.sf files have NM_ (vaidated) IDs. So there is mismatch problem. I want to know the reason

[ATpoint]

Check for the intersect between the transcript names. They should be (largely) identical.

[Sacit]

Could you please see this comment? I am facing the same:

How do I match my transcript ID's from NCBI to the corresponding gene ID's to enable tximport into R?

Answer 1

[Sacit]

When I set ignoreTxVersion = FALSE, it seems to work

By default, it is FALSE