Circrotate

Circovirus Rotator

This software takes circovirus genomes, finds replication and capsid proteins through Prokka, and rotates the genome so the replication origin is the first feature in the linearized record.

Dependencies:

To run properly, this pipeline needs:

Recommended versions of these dependencies are in the environment.yml file, for use with Conda virtual environments.

Options:

Option	Use	Default
--in	Specify the input file(s) in fasta format	You must specify an input
--out	Specify the ouput folder name	"pipe_out"
--motif	Specify the JASPAR sites file to use	circovirus replication origins in "testData/rep_orig_circo.sites"
--prokkaOpts	Extra prokka options. Must be wrapped in quotes	""
--seqBatch	Specify the number of sequences processed per thread	1 sequence per thread
--genBank	Flag for genbank format files	false

Output files:

(named output)/
- prokka-annotations/
  - One folder for each annotation format, each containing annotation files for each batch of processed sequences.
  - spf/
    - folder containing the sed patterns later used to restore the fasta descriptions to the output files.
- sequences/
  - fasta/ organized by batch
    - the fasta files of the rotated sequences.
  - gff/ organized by batch
    - the gff annotation files from Prokka.
  - txt/ organized by batch
    - the txt files containing the lists of names of sequences that could not be rotated.
  - "all_sequences.fasta" contains all the rotated sequences in one fasta file.
  - "all_fails.txt" contains all the names of sequences that could not be rotated.
  - "all_annotations.gff" contains all the prokka annotations of rotated sequences.
  - "all_results.csv" contains a summary of the results.
  - "annotatedFastas.gff" contains all the annotations in gff format, plus all the sequences in a ##FASTA section.

Result CSVs:

CSVs created as reports contain the following information:

Name: Sequence ID.
Success: Successful rotation of the sequence.
Shift: Position the record was shifted to, relative to its old base numbering.
Reversed: Was the sequence reverse complemented.
Genbank retrieved: Did this sequence use a genbank file for its output annotations.
Step failed: For failed rotations, the step that failed. Either "annotation" if Prokka didn't find the rep or cap genes, or "motif" if the rep origin sequence couldn't be located.
Possible rep origins: Locations of other possible rep origins, relative to the originl record's numbering.

Test command

This pipeline comes with some circovirus genomes from NCBI as test material, as well as 6 replication origins to search for: NAGTATTAC; YATTATTAC. You may wish to make your own sites file with all replication origins mentioned by ICTV's Circovirus fact sheet: NANTATTAC.

Test commands:

Normal usage (fasta files):

nextflow run Circrotate --in Circrotate/testData/testFasta.fasta

This command can be run by using -profile test.

Usage with genbank (*.gb) files:

nextflow run Circrotate --in Circrotate/testData/circo.gb --genbank

This command can be run by using -profile testgb.

All sequences in both test sets should be rotated to the replication origin.

Notes:

The program tries to use the start of the rep or cap gene as an anchor for the expected replication origin, in case there is a siimlar sequence elsewhere in the genome. However, this is not perfect, and the following errors can occur:

No annotations: If Prokka doesn't find either rep or cap genes, the program discards the sequence and doesn't attempt to rotate it.
No motif: If the rep origin motif can't be located anywhere, the sequence is discarded.
Incorrect sequence: Since the rep origin location is based on proximity to one gene alone, it is possible that a similar sequence can exist in the gene closer to the start than the true replication origin, meaning the genome would be rotated to the sequence in the gene. It is also possible that the only seuqence recognised is in the incorrect place.

All sequences have to be re-annotated to get a consistent naming convention for the gene products. Using (inconsistent) manual annotations will probably take some regex tricks that the python script currently isn't set up to handle.

Versioning:

versioning: X.Y.Z

X: Major release version

Anything that adds an analysis step.
Things like adding a BLAST search, adding options for genome annotation programs, etc.
Changes the core functionality of the program.

Y: Minor release version

Changing the interface or adding smaller features, things that don't affect the core functionality of the program.
Giving a command line option to add Prokka options, changing output folder strucutre, etc.
Optionally different, but produces the same type of output and accepts the same inputs.

Z: Bugfix version

For when I make mistakes and have to push another version to change something that otherwise breaks the program.
Also used for back-end improvements that don't actually affect the usage or output.

Licensing:

This project is licensed under the MIT license. See license.txt for details.

TMRHarrison/Circrotate