Fail to plot jaccard and region_type 'Axes' object is not subscriptable error leads to malformed file output

[byee4]

I think there is a bug near or at this line:

Mudskipper/scripts/analyze_DMM.py

Line 536 in 9a44786

print('Fail to plot jaccard and region_type', e)

It seems to be causing incomplete file creation for some RBPs, and I'm not sure yet what is different about these RBPs than others, but it ends up creating a partial file (attachment)

errored_windows.txt

[byee4]

Just kidding. I think changing "ax" to "axes" fixes the error, but the bad files still remain

Mudskipper/scripts/analyze_DMM.py

Lines 529 to 530 in 9a44786

	ax[0].legend(bbox_to_anchor=(1,1))
	ax[1].legend(bbox_to_anchor=(1,1))

[byee4]

For some reason I don't understand, these RBPs fail the masking step and thus do not have any windows to output (enriched_windows is an empty DataFrame). Do you think adding a shape[0] > 0 check would simply fix the error?

As in, replace these lines:

Mudskipper/scripts/analyze_DMM.py

Lines 535 to 556 in 9a44786

	except Exception as e:
	print('Fail to plot jaccard and region_type', e)
	# ====== Generate Output ======
	# --- output by individual RBP ---
	columns = ['chrom', 'start', 'end', 'name', 'score',
	'strand', 'feature_id', 'feature_bin', 'feature_type_top',
	'feature_types', 'gene_name', 'gene_id', 'transcript_ids',
	'gene_type_top', 'transcript_type_top', 'gene_types',
	'transcript_types','logLR', 'BF_assignment',]
	# output per RBP enrich windows
	for c in anno.columns:
	enriched_windows = data_bf_dmm_masked.loc[data_bf_dmm_masked[c], columns]
	enriched_windows['p_bar']=p_bar[c]
	enriched_windows['p_raw']=p_raw[c]
	enriched_windows['fc_raw']=fc_raw[c]
	enriched_windows['fc_bar']=fc_bar[c]
	enriched_windows['p_bar_std']=std_bar[c]
	enriched_windows.to_csv(basedir / 'DMM'/ f'{c}.enriched_windows.tsv', sep = '\t')
	print(f'found {c} enriched windows:', enriched_windows.shape)

with this?

    except Exception as e:
        print('Fail to plot jaccard and region_type', e)

    # ====== Generate Output ======
    # --- output by individual RBP ---
    columns = ['chrom', 'start', 'end', 'name', 'score',
    'strand', 'feature_id', 'feature_bin', 'feature_type_top',
    'feature_types', 'gene_name', 'gene_id', 'transcript_ids',
    'gene_type_top', 'transcript_type_top', 'gene_types',
    'transcript_types','logLR', 'BF_assignment',]
    # output per RBP enrich windows

    for c in anno.columns:
        enriched_windows = data_bf_dmm_masked.loc[data_bf_dmm_masked[c], columns]
        if enriched_windows.shape[0] > 0:
            enriched_windows['p_bar']=p_bar[c]
            enriched_windows['p_raw']=p_raw[c]
            enriched_windows['fc_raw']=fc_raw[c]
            enriched_windows['fc_bar']=fc_bar[c]
            enriched_windows['p_bar_std']=std_bar[c]
        enriched_windows.to_csv(basedir / 'DMM'/ f'{c}.enriched_windows.tsv', sep = '\t')
        print(f'found {c} enriched windows:', enriched_windows.shape)