'Hjurp',
'Kpna2',
'Kif23',
'Cks2',
'Dtl',
'Top2a',
'Bub1',
'Arl6ip1',
'Dlgap5',
'Ube2s',
'Nuf2',
'Hmmr',
'Cdc20',
'Birc5',
]
dscc = ds.query_features_by_name(genes_cc2, inplace=False)
dscc.counts.log(inplace=True)
vsu = dscc.dimensionality.umap()
fig, axs = plt.subplots(2, 4, sharex=True, sharey=True, figsize=(9, 5))
genes_plot = ['log_virus_reads_per_million', 'Ccne2', 'Ccnd1', 'Ccnb1']
for icol, gene in enumerate(genes_plot):
ax = axs[0][icol]
dscc.plot.scatter_reduced_samples(vsu,
color_by=gene,
alpha=0.4,
ax=ax,
s=10)
ax.set_title(gene)
ax.set_axis_off()
ax = axs[1][icol]
if gene in dscc.counts.index:
dss = ds.split('clusterN_SNV_{:}'.format(ic))
comp = dss[True].compare(dss[False])
# FIXME: maybe look symmetrically for up- and downregulated
comp['diff'] = dss[True].counts.mean(axis=1) - dss[False].counts.mean(
axis=1)
genesi = comp.loc[comp['diff'] > 0,
'P-value'].nsmallest(n=5).index.values
genes[ic] = genesi
genes_all = np.unique(np.concatenate(list(genes.values())))
with open('../data/genes_diff_expressed_clustersSNV.tsv', 'wt') as f:
f.write('\t'.join(genes_all))
dsv = ds.query_features_by_name(genes_all)
# Plot distributions
fig, axs = plt.subplots(3, 10, figsize=(17, 7), sharex=True, sharey=True)
axs = axs.ravel()
for ax, gene in zip(axs, genes_all):
df = np.log10(0.1 + dsv.counts.loc[[gene]].T)
df['clusterN'] = dsv.samplesheet['clusterN_SNV']
sns.boxplot(
data=df,
y=gene,
x='clusterN',
ax=ax,
order=clusters,
)
ax.grid(axis='y')
# Calculate transciptome distances
ds = Dataset(
samplesheet='dengue',
counts_table='dengue',
featuresheet='humanGC38',
)
ds.samplesheet['cluster_SNV'] = clusters
ds.counts.normalize(inplace=True)
ds.rename(axis='features', column='GeneName', inplace=True)
ds.feature_selection.unique(inplace=True)
# Restrict to differentially expresse genes
with open('../data/genes_diff_expressed_clustersSNV.tsv', 'rt') as f:
genes = f.read().split('\t')
dsd = ds.query_features_by_name(genes)
dsd.counts.log(inplace=True)
dsp = dsd.split('cluster_SNV')
dclut = {}
for i1, c1 in enumerate(clustersu):
ge1 = dsp[c1].counts.values.T
for c2 in clustersu[:i1+1]:
print(c1, c2)
ge2 = dsp[c2].counts.values.T
if c1 != c2:
d = cdist(ge1, ge2).ravel()
else:
d = squareform(cdist(ge1, ge2))
dclut[frozenset([c1, c2])] = d
