How important are the genes to explain the outcome - the asymmetric Shapley value as an honest importance metric for high-dimensional features

Mark A. van de Wiel; Jeroen Goedhart; Martin Jullum; Kjersti Aas

stat.MLMar 5, 2026

How important are the genes to explain the outcome - the asymmetric Shapley value as an honest importance metric for high-dimensional features

Authors: Mark A. van de Wiel, Jeroen Goedhart, Martin Jullum, Kjersti Aas

Organizations: Dept of Epidemiology and Data Science, Amsterdam Public Health research institute, Amsterdam University Medical Centers, Amsterdam, The Netherlands · Norwegian Computing Center, Oslo, Norway

Abstract

In clinical prediction settings the importance of a high-dimensional feature like genomics is often assessed by evaluating the change in predictive performance when adding it to a set of traditional clinical variables. This approach is questionable, because it does not account for collinearity nor known directionality of dependencies between variables. We suggest to use asymmetric Shapley values as a more suitable alternative to quantify feature importance in the context of a mixed-dimensional prediction model. We focus on a setting that is particularly relevant in clinical prediction: disease state as a mediating variable for genomic effects, with additional confounders for which the direction of effects may be unknown. We derive efficient algorithms to compute local and global asymmetric Shapley values for this setting. The former are shown to be very useful for inference, whereas the latter provide interpretation by decomposing any predictive performance metric into contributions of the features. Throughout, we illustrate our framework by a leading example: the prediction of progression-free survival for colorectal cancer patients.

Figures & tables

Shapley value ∗	Local feature importance (explaining individual prediction)
SAGE	Global feature importance (explaining model performance)
Symmetric	No directional effects
Asymmetric	Directional and undirectional effects
SHAP	Estimation by marginalizing over features not in the subset
Refitting	Estimation by refitting models on subsets of features
Marginal approach	SHAP assuming independence for marginalization

Table 1: Concepts related to Shapley values.

Figure 1: Model for information flow for

G

: gene expression profile,

C

: confounders,

D

: disease state,

Y

: outcome. Right arrows depict unidirectional effects, bidirectional arrows depict correlations. Grey arrows indicate prediction, black arrows dependency between variables. Modules may contain multiple features.

	C-index	tAUC
blockForest	0.754	0.789
fusedTree	0.747	0.777
ridge0	0.710	0.715
blockForest (reduced)	0.731	0.737

Table 2: Predictive performances

	Sym	Asym	$\text{LLR}_{\text{Sym}}$	$\text{LLR}_{\text{Asym}}$	$p_{\text{Sym}}$	$p_{\text{Asym}}$
intercept	0.500	0.500
Genes ( $G$ )	0.130	0.173	12.39	14.56	1.72e-05	2.12e-06
– $G$ High ( $G_{1}$ )	0.076	0.082	7.87	7.14	7.27e-05	1.57e-04
– $G$ Low ( $G_{2}+G_{3}$ )	0.053	0.091	2.28	4.91	0.102	7.39e-03
— CMS ( $G_{2}$ )	0.028	0.043	1.59	2.17	0.075	0.037
— Pred $D$ ( $G_{3}$ )	0.025	0.048	0.43	2.36	0.356	0.030

Table 3: SAGE decomposition of C-index of blockForest model for symmetric and asymmetric version. Plus (partial) likelihood-ratio (log; LLR) and

p

-values based on the likelihood-ratio test using Shapley values on log cumulative hazard scale.

Figure 2: Mediation effect. Shapley values for Genes

G

(y-axis) against true disease state

D

(x-axis) for the asymmetric (left box) and symmetric (right box) version.

	$G$	$D$	$C_{1}$	$C_{2}$	$C_{3}$	$C_{4}$	$C_{5}$	$C_{6}$	$C_{7}$	$C_{8}$	$C_{9}$	$C_{10}$	$C_{11}$	$C_{12}$
$\hat{\phi}^{+}_{h}$	1.01	1.00	0.96	0.98	1.01	1.02	1.01	1.01	1.00	1.01	1.00	0.99	1.03	0.98
$\hat{\phi}^{-}_{h}$	0.99	1.01	1.00	1.00	1.02	1.00	1.00	1.00	1.01	0.97	1.01	1.00	0.97	0.99

Table 4: Estimated asymmetric Shapley values for constant contribution function. Estimation based on importance sampling with

B=10,000

sampled subsets

Variable	Asymmetric	Symmetric	Marginal
$G$	$G\left(\beta_{1}+\beta_{2}\gamma\right)$	$\beta_{1}G+\frac{\gamma}{2}\!\left(\beta_{2}G-\beta_{1}D\right)$	$\beta_{1}G$
$D$	$\beta_{2}\!\left(D-\gamma G\right)$	$\beta_{2}D+\frac{\gamma}{2}\!\left(\beta_{1}D-\beta_{2}G\right)$	$\beta_{2}D$

Table 5: Asymmetric, symmetric, and marginal Shapley values (

\phi

) for the toy example described by ( 7 ).

Figure 3: Analytical Shapley values for variables

G

(bottom panels) and

D

(top panels) as a function of the variable values. Both the symmetric (black) and the asymmetric (orange) versions are depicted. Results are shown for varying correlation strengths

\rho

. Remaining parameter settings:

\beta_{1}=\beta_{2}=1.0

\beta_{3}=5.0

, and

\gamma=0.8

. The plot is produced using the R -package ggplot2 ( Wickham, 2016 ) .

Figure 4: Shapley scatter plots for 200 test observations for the low-dimensional simulation. Variable on the x-axis, its Shapley value on y-axis. All Shapley values are estimated using SHAP, i.e. marginalisation.

	Sym-marg	Sym-cond	Asym
$C_{1}$	0.11	0.11	0.11
$C_{2}$	0.00	0.02	0.04
$D$	0.66	0.50	0.26
$G$	0.11	0.25	0.48
Total	0.88	0.88	0.88

Table 6: SAGE-

R^{2}

on test observations for the low-dimensional simulation.

Figure 5: Asymmetric Shapley values based on refitting. Scatter plots for 200 test observations for the low-dimensional simulation. Variable on the x-axis, its Shapley value on y-axis.

	fusedTree		ridge0
	Sym	Asym	Sym	Asym
Intercept	0.500	0.500	0.500	0.500
Genes ( $G$ )	0.122	0.170	0.075	0.143
Disease state ( $D$ )	0.121	0.071	0.118	0.058
Gender ( $C_{1}$ )	0.003	0.005	0.028	0.008
Age ( $C_{2}$ )	0.005	0.007	0.003	0.005

Table 7: SAGE decomposition of C-index of fusedTree and ridge0 models for asymmetric and symmetric setting

Figure 6: Symmetric and asymmetric Shapley values for

G

and

D

for the blockForest model. Shapley values were computed on the log cumulative hazard scale

	Sym	Asym	$p_{\text{Sym}}^{\text{NP}}$	$p_{\text{Asym}}^{\text{NP}}$
intercept	0.500	0.500
Genes ( $G$ )	0.130	0.173	4.39e-3	5.57e-4
– $G$ High ( $G_{1}$ )	0.076	0.082	0.058	0.039
– $G$ Low ( $G_{2}+G_{3}$ )	0.053	0.091	0.502	0.014
— CMS ( $G_{1}$ )	0.028	0.043	0.398	0.054
— Pred $D$ ( $G_{2}$ )	0.025	0.048	0.717	0.245

Table 8: SAGE decomposition of C-index of blockForest model for symmetric and asymmetric version. Plus significance based on the nonparametric (NP) conditional independence test using Shapley values on log cumulative hazard scale.

Explore similar work

cc-Shapley: Measuring Multivariate Feature Importance Needs Causal Context
Feb 23, 2026Jörg Martin, Stefan HaufeFeature Attribution Shapley Value Attribution
Physikalisch-Technische Bundesanstalt, Berlin, Germany · Technische Universität Berlin, Berlin, Germany · Charité – Universitätsmedizin Berlin, Berlin, Germany
MinShap: A Modified Shapley Value Approach for Feature Selection
Apr 16, 2026Chenghui Zheng, Garvesh RaskuttiShapley Value Attribution Feature Selection
Department of Statistics, University of Wisconsin - Madison, Madison, WI 53706
Explaining a probabilistic prediction on the simplex with Shapley compositions
Aug 2, 2024Paul-Gauthier Noé, Miquel Perelló-Nieto, Jean-François Bonastre +1Shapley Value Attribution
Laboratoire Informatique d’Avignon, Avignon Universit´e, France · University of Bristol, United Kingdom

Feb 23, 2026J/K move · Enter open · S save

cc-Shapley: Measuring Multivariate Feature Importance Needs Causal Context

Jörg Martin, Stefan Haufe

Physikalisch-Technische Bundesanstalt, Berlin, Germany · Technische Universität Berlin, Berlin, Germany · Charité – Universitätsmedizin Berlin, Berlin, Germany

Feature Attribution Shapley Value Attribution Causal Attribution Causal Inference

Open paper AskPDF arXiv

Explainable artificial intelligence promises to yield insights into relevant features, thereby enabling humans to examine and scrutinize machine learning models or even facilitating scientific discovery. Considering the widespread technique of Shapley values, we find that purely data-driven operationalization of multivariate feature importance is unsuitable for such purposes. Even for simple problems with two features, spurious associations due to collider bias and suppression arise from considering one feature only in the observational context of the other, which can lead to misinterpretations. Causal knowledge about the data-generating process is required to identify and correct such misleading feature attributions. We propose cc-Shapley (causal context Shapley), an interventional modification of conventional observational Shapley values leveraging knowledge of the data's causal structure, thereby analyzing the relevance of a feature in the causal context of the remaining features. We show theoretically that this eradicates spurious association induced by collider bias. We compare the behavior of Shapley and cc-Shapley values on various, synthetic, and real-world datasets. We observe nullification or reversal of associations compared to univariate feature importance when moving from observational to cc-Shapley.

Abstract

Explore similar work

cc-Shapley: Measuring Multivariate Feature Importance Needs Causal Context

MinShap: A Modified Shapley Value Approach for Feature Selection

Explaining a probabilistic prediction on the simplex with Shapley compositions