Circular contigs

Identify plasmid sequences which can be made circular.

changed milestone to %Stretch goal

added Analysis Somewhere over the rainbow labels

changed title from Plasflow: cricular plasmids to Circular plasmids

changed the description

BLAST-based approach: Use rules from /mnt/lscratch/users/sbusi/ONT/cedric_ont_basecalling/2019_GDB/mock/circular/snakefile

Would only work if the ends do really overlap which may be not true depending on the used assembler.

• TODO: we need to account for where the matches occur. For eg. we're looking for matches at the beginning of one string against that at the end of the other. Not sure if there's an easy way to automate this yet

• repo: (a) (p)erfect (c)ircle?
• script: identifyCircularContigs

Currently not used

removed Somewhere over the rainbow label

changed milestone to %Analyses

circlator

• to identify circular sequences
• assembling all reads that map to contig ends and comparing the resulting contigs with the input assembly
• primarily for long reads

Currently not using for all assemblies because would need different parameters/commands for different assembly types.

We could use it for long-read assembly only to compare the results to the BLAST-based approach.

mentioned in commit 1ebb366e

mentioned in commit 334acc25

marked this issue as related to #4 (closed)

changed title from Circular plasmids to Circular contigs

Closing for now

• blast-based approach does not work if there are no overlapping ends
  • depends of the assembler
• other approaches rely on read alignment and/or assembly graph
• nice to have but not the focus of this study

closed

mentioned in issue #4 (closed)