Dynamical Causality Under Latent Confounders for Biological Network Reconstruction
- 28 January 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 48 (6) , 6703-6719
Abstract
Causal interaction inference is prone to spurious causal interactions, due to the substantial confounders in a biological system. While many existing methods attempt to address misidentification challenges, there remains a notable lack of effective methods to infer causal interaction under latent/unobserved confounders. In this work, we propose a method to overcome such challenges to infer dynamical causality under invisible confounders (CIC) and further reconstruct the latent confounders from time-series data by developing an orthogonal decomposition theorem in a delay embedding space. This theoretical foundation ensures the causal detection for any high-dimensional system even with only two observed variables under many latent confounders, which is a long-standing problem in the field. In addition to the latent confounder problem, such a decomposition makes the coupled variables separable in the embedding space, thus also solving the non-separability problem of causal inference. Extensive validation of the CIC method is carried out using various real datasets, which all demonstrates its effectiveness to reconstruct real biological networks and unobserved confounders.Keywords
Funding Information
- National Natural Science Foundation of China (T2350003)
- National Natural Science Foundation of China (T2341007)
- National Natural Science Foundation of China (62473312)
- National Natural Science Foundation of China (62173271)
- National Natural Science Foundation of China (12131020)
- National Natural Science Foundation of China (42450084)
- National Natural Science Foundation of China (42450135)
- National Natural Science Foundation of China (12426310)
- National Natural Science Foundation of China (12301620)
- National Natural Science Foundation of China (42450192)
- National Natural Science Foundation of China (T2542018)
- National Natural Science Foundation of China (12401661)
- National Key R&D Program of China (2025YFC3409300)
- National Key R&D Program of China (2025YFF1207900)
- National Key R&D Program of China (2022YFA1004800)
- Science and Technology Commission of Shanghai Municipality (23JS1401300)
- Shanghai Municipal Data Bureau (202401065)
- AI for Science Foundation of Fudan University (FudanX24AI041)
- Zhejiang Province Vanguard Goose-Leading Initiative (2025C01114)
- Shenzhen Medical Research Fund (E250200621)
- JST Moonshot R&D (JPMJMS2021)
- TianFu JinCheng Laboratory (TFJCPI20260001)
This publication has 63 references indexed in Scilit:
- Detecting dynamical causality by intersection cardinal concavityFundamental Research, 2025
- Deep recurrent modelling of Granger causality with latent confoundingExpert Systems with Applications, 2022
- Inferring causal molecular networks: empirical assessment through a community-based effortNature Methods, 2016
- Global association of air pollution and heart failure: a systematic review and meta-analysisThe Lancet, 2013
- Main Air Pollutants and Myocardial InfarctionJAMA, 2012
- Accurate modeling of confounding variation in eQTL studies leads to a great increase in power to detect trans-regulatory effectsNature Precedings, 2011
- Chaos in a long-term experiment with a plankton communityNature, 2008
- System-level identification of transcriptional circuits underlying mammalian circadian clocksNature Genetics, 2005
- A robust method for detecting interdependences: application to intracranially recorded EEGPhysica D: Nonlinear Phenomena, 1999
- Estimating causal effects of treatments in randomized and nonrandomized studies.Journal of Educational Psychology, 1974