File size: 1,421 Bytes
925ee3b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 | """Variance decomposition into transcriptional and post-transcriptional fractions."""
from __future__ import annotations
import numpy as np
from anndata import AnnData
from .._constants import (
SMOOTHED_UNSPLICED,
GAMMA,
TF_SCORE,
PTF_SCORE,
)
from .._utils import get_layer, require_layers, log_params
def variance_decomposition(adata: AnnData) -> None:
"""Decompose gene expression variance into TF and PTF components.
In log space:
- ``TF_g = Var(log(u_g)) / (Var(log(u_g)) + Var(log(gamma_g)))``
- ``PTF_g = 1 - TF_g``
Results are stored in ``adata.var['tf_score']`` and ``adata.var['ptf_score']``.
Parameters
----------
adata
Annotated data matrix with smoothed unspliced layer and gamma.
"""
require_layers(adata, SMOOTHED_UNSPLICED, GAMMA)
u = get_layer(adata, SMOOTHED_UNSPLICED)
gamma = get_layer(adata, GAMMA)
# Log-transform with pseudocount
log_u = np.log1p(u)
log_gamma = np.log1p(gamma)
var_u = np.var(log_u, axis=0)
var_gamma = np.var(log_gamma, axis=0)
total_var = var_u + var_gamma
# Avoid division by zero
total_var = np.clip(total_var, 1e-10, None)
tf_score = var_u / total_var
ptf_score = var_gamma / total_var
adata.var[TF_SCORE] = tf_score.astype(np.float32)
adata.var[PTF_SCORE] = ptf_score.astype(np.float32)
log_params(adata, "variance_decomposition", {})
|