def build_from_mol_counter(molecule_counter, subsample_rate=1.0,
subsample_result=None):
""" Construct a GeneBCMatrices object from a MoleculeCounter.
Args: subsample_result (dict) - Return some metrics results into this dict. """
# Reconstruct all barcode sequences in the original matrices
barcode_whitelist = cr_utils.load_barcode_whitelist(molecule_counter.get_barcode_whitelist())
barcode_length = molecule_counter.get_barcode_length() or len(barcode_whitelist[0])
gem_groups = molecule_counter.get_gem_groups()
barcode_seqs = cr_utils.format_barcode_seqs(barcode_whitelist, gem_groups)
# Reconstruct Gene tuples from the molecule info ref columns
gene_ids = molecule_counter.get_ref_column('gene_ids')
genome_ids = molecule_counter.get_ref_column('genome_ids')
gene_names = molecule_counter.get_ref_column('gene_names')
gene_tuples = [cr_constants.Gene(gid, gname, None, None, None) for (gid, gname) in itertools.izip(gene_ids, gene_names)]
genes = cr_utils.split_genes_by_genomes(gene_tuples, genome_ids)
matrices = GeneBCMatrices(genome_ids, genes, barcode_seqs)
# Track results of subsampling
reads = 0
for mol in molecule_counter.get_molecule_iter(barcode_length, subsample_rate=subsample_rate):
matrices.add(mol.genome, mol.gene_id, mol.barcode)
reads += mol.reads
if subsample_result is not None:
subsample_result['mapped_reads'] = reads
return matrices
def main(args, outs):
in_bam = tk_bam.create_bam_infile(args.chunk_input)
chroms = in_bam.references
barcode_whitelist = cr_utils.load_barcode_whitelist(args.barcode_whitelist)
barcode_summary = cr_utils.load_barcode_summary(
args.barcode_summary) if not barcode_whitelist else None
gene_index = cr_reference.GeneIndex.load_pickle(
cr_utils.get_reference_genes_index(args.reference_path))
reporter = cr_report.Reporter(reference_path=args.reference_path,
high_conf_mapq=cr_utils.get_high_conf_mapq(
args.align),
gene_index=gene_index,
chroms=chroms,
barcode_whitelist=barcode_whitelist,
barcode_summary=barcode_summary,
gem_groups=args.gem_groups)
if barcode_whitelist:
barcode_seqs = cr_utils.format_barcode_seqs(barcode_whitelist,
args.gem_groups)
else:
barcode_seqs = barcode_summary
genomes = cr_utils.get_reference_genomes(args.reference_path)
genes = cr_utils.split_genes_by_genomes(gene_index.get_genes(), genomes)
matrices = cr_matrix.GeneBCMatrices(genomes, genes, barcode_seqs)
for read in in_bam:
is_conf_mapped_deduped, genome, gene_id, bc = reporter.count_genes_bam_cb(
read, use_umis=cr_chem.has_umis(args.chemistry_def))
if is_conf_mapped_deduped:
matrices.add(genome, gene_id, bc)
in_bam.close()
matrices.save_h5(outs.matrices_h5)
reporter.save(outs.chunked_reporter)