Average length of mRNA on NCBI annotation seems too high

Question

[yh1126]

I noticed that average length of all mRNA transcripts from human (GRCh38) annotation was 90043.2 bp, which seems unusually high. The average length of mRNA should be 3-4,000 bp as I am aware.

Is there an explanation for this?

I calculated the average length from all entry records on GRCh38 RefSeq annotation for which the column 2 (feature type) information was “mRNA”. Below is the command that I used.

awk -v FS="\t" '$1!~"^#" && $3=="mRNA"{sum+=$5-$4;n++}END{print sum/n}' GCF_000001405.40_GRCh38.p14_genomic.gff

The “awk” that I’m using is mawk 1.3.4.

Answer 1

[lieven.sterck]

Seems like you (wrongly) calculated the average mRNA length on the genome (== including introns ) and thus not only the summed length of all the exons on a gene (== biological mRNA, == what you want or assumed you calculated)

You should sum&average the accumulated length of all the exon entries per mRNA in your file