De novo assembly average length

Question

[aatif1903]

Hello everyone. I have did de novo assembly by using CLC Bio. I have an average length of transcripts about 119bp. Is this length enough for correct annotation?

[WouterDeCoster]

Average is a bad statistic (of most things). Try to plot the distribution, maybe the average get's dragged down by unassembled short fragments.

On average, a human has slightly less than one testicle. I think that we can agree that that is not representative of the true distribution. On average, there are two popes per square kilometer in the Vatican.

Answer 1

[Brian Bushnell]

Average length is not a useful statistic in assembly. L50 (the length of the at which 50% of the assembled bases are in contigs of at least that length) is much more useful.

Generally, 119 bp contigs are useless for annotation; and where your contigs are shorter than your reads, assembly failed.