metaplasmidSPAdes reconstructs identical plasmids

[barbaracania]

Hello!
I noticed that metaplasmidSPAdes reconstructed the exact same plasmid twice (I checked the sequences) using a different cutoff. The names of the identical contigs are: NODE_1_length_65404_cov_23.803361_cutoff_10_type_circular and NODE_1_length_65404_cov_23.803361_cutoff_15_type_circular. Is it an artifact or is there a reason for such behavior?
Best,
Barbara

[asl]

Tagging @Dmitry-Antipov – they author of metaplasmidSPAdes

[Dmitry-Antipov]

Hi.
Sorry for the delay. This looks like a bug. Can you share the data? it could simplify the fixing.

If not, can you check whether these sequences are identical or have some minor differences (i.e. circular shift or SNP) and attach spades.log file?

[Dmitry-Antipov]

Thank you
Seems that we still need the reads to understand why this may happen. Can you share them with us (i.e through google drive, to [email protected])?

[Dmitry-Antipov]

Sorry for bothering you again, but still - could you send the data to us to reproduce the issue?

[barbaracania]

No problem at all, I am sorry myself that I did not contact you earlier with some info on when I would be able to share the data. I have uploaded the reads to google drive now and made them available for you. As for the parameters used, I ran spades only with the --metaplasmid flag, everything else was set to default.

[Dmitry-Antipov]

Got it, thank you!

[Dmitry-Antipov]

Thank you, now the reason of this problem is clear.
These duplication is an artifact (a bug, actually) ), no real biological reasons.

Hope to fix this till next release.

[asl]

The issue was addressed in https://github.com/ablab/spades/releases/tag/v3.15.5