diffsvg.tl.cluster_curves#
- diffsvg.tl.cluster_curves(adata, by='average', n_clusters=5, genes=None, scale=False, coefs_key='dsvg_coefs', profiles_key='dsvg_profiles', key_added=None, random_state=0, n_init=100, inplace=True)#
Group genes by the shape of their fitted curves, with k-means on the spline coefficients.
- Parameters:
- adata
Annotated data matrix that
diffsvg.tl.fit_model(ordiffsvg.tl.fit_profiles, forby="profile") wrote into.- by default:
'average' Which curves to compare:
"average":beta + beta_1 / 2, the combined curve: the mean of the two condition curves, weighting the conditions equally. Notbetaalone, which with the 0/1 condition coding is the curve of condition 0."interaction": the interaction termbeta_1, i.e. the shape of the difference curvef1 - f0. Not the level difference, which is a separate coefficient (categorical_beta)."profile":betaofdiffsvg.tl.fit_profiles, the one curve of a fit that ignores the condition. Without slide effects.
- n_clusters default:
5 Number of clusters (
k).- genes default:
None Genes to cluster, by default every fitted one.
- scale default:
False Scale each gene’s coefficients to unit length, so that genes group by shape alone, however strong the trend. By default (
False) the amplitude counts, so genes without a trend share one flat cluster. Scaling blows the noise of such genes up to the size of real trends, so combine it withgenes, e.g. the genes with a clear trend.- coefs_key default:
'dsvg_coefs' Prefixes passed to
diffsvg.tl.fit_model/diffsvg.tl.fit_profiles.- profiles_key default:
'dsvg_profiles' Prefixes passed to
diffsvg.tl.fit_model/diffsvg.tl.fit_profiles.- key_added default:
None Where to write the result. Defaults to
f"dsvg_cluster_{by}".- random_state default:
0 Passed to
sklearn.cluster.KMeans. Many restarts cost next to nothing with one row per gene, and they make the result depend far less on the seed: on an 88-gene panel 10 gave up to 25 different partitions across 30 seeds, 100 mostly one.- n_init default:
100 Passed to
sklearn.cluster.KMeans. Many restarts cost next to nothing with one row per gene, and they make the result depend far less on the seed: on an 88-gene panel 10 gave up to 25 different partitions across 30 seeds, 100 mostly one.- inplace default:
True If
True(default), writeadata.var[key_added](categorical,NaNfor genes not clustered) andadata.uns[key_added](by,scale,n_clusters,silhouette,inertia,centers) and returnNone. Otherwise return the labels as apandas.Series.
- Returns:
None or pandas.Series Labels are
"0","1", …, the largest cluster first.
Examples
>>> dvg.tl.cluster_curves(adata, by="average", n_clusters=6) >>> dvg.pl.plot_cluster_curves(adata, "dsvg_cluster_average")