This should probably be broken up into a subworkflow where ANALYZE_KMERS would be

[chasemc]

This should probably be broken up into a subworkflow where ANALYZE_KMERS would be

workflow ANALYZE_KMERS {
  take:
    fasta
  main:
    COUNT_KMERS(fasta)
    NORMALIZE_KMERS(COUNT_KMERS.out)
    EMBED_KMERS(NORMALIZE_KMERS.out)
  emit:
    counts = COUNT_KMERS.out
    normalized = NORMALIZE_KMERS.out
    embedded = EMBED_KMERS.out
}

Originally posted by @WiscEvan in #157 (comment)

[chasemc]

relevant current code:
https://github.com/KwanLab/Autometa/blob/dev/modules/local/analyze_kmers.nf

[chasemc]

@WiscEvan I can't see that comment anymore, just wanted to confirm this still what we want to do?
-> Basically split the single Python endpoint into 3 endpoints and then make nextflow processes for those 3, and wrap those 3 nextflow processes into a sub workflow

@Sidduppal You may be comfortable enough with NF now to tackle this one if you wanted.

Relevant code:

Autometa/autometa/common/kmers.py

Lines 718 to 743 in f372b82

	if os.path.exists(args.kmers) and not args.force:
	df = pd.read_csv(args.kmers, sep="\t", index_col="contig")
	else:
	df = count(
	assembly=args.fasta,
	size=args.size,
	out=args.kmers,
	force=args.force,
	cpus=args.cpus,
	)
	if args.norm_output:
	df = normalize(
	df=df, method=args.norm_method, out=args.norm_output, force=args.force
	)
	if args.embedding_output:
	embedded_df = embed(
	kmers=df,
	out=args.embedding_output,
	force=args.force,
	method=args.embedding_method,
	embed_dimensions=args.embedding_dimensions,
	pca_dimensions=args.pca_dimensions,
	seed=args.seed,
	)

https://github.com/KwanLab/Autometa/blob/dev/modules/local/analyze_kmers.nf

[evanroyrees]

We should be able to implement this in nextflow without changing the autometa-kmers entrypoint..

e.g.

# process: count-kmers
autometa-kmers --fasta $fasta --kmers $counts --size $size --cpus $cpus

# process:  normalize-kmers
autometa-kmers --fasta $fasta --kmers $counts --size $size --norm-method $method --norm-output $output

# process: embed-kmers
autometa-kmers \
    --fasta $fasta \
    --kmers $counts \
    --norm-output $output \
    --embedding-method $method \
    --seed $embed_seed \
    --embedding-dimensions $embed_dims \
    --embedding-output $output