Challenges and Opportunities with Causal Discovery Algorithms: Application to Alzheimer's Pathophysiology

Xinpeng Shen; Sisi Ma; Prashanthi Vemuri; Gyorgy Simon; Alzheimer’s Disease Neuroimaging Initiative

doi:10.1038/s41598-020-59669-x

Back

Challenges and Opportunities with Causal Discovery Algorithms: Application to Alzheimer's Pathophysiology

Journal article

Open access

Peer reviewed

Challenges and Opportunities with Causal Discovery Algorithms: Application to Alzheimer's Pathophysiology

Xinpeng Shen, Sisi Ma, Prashanthi Vemuri, Gyorgy Simon and Alzheimer’s Disease Neuroimaging Initiative

Scientific reports, Vol.10(1), pp.2975-2975

02/19/2020

DOI: 10.1038/s41598-020-59669-x

PMCID: PMC7031278

PMID: 32076020

Files and links (1)

url

https://doi.org/10.1038/s41598-020-59669-xView

Published (Version of record) Open Access

Abstract

Causal Structure Discovery (CSD) is the problem of identifying causal relationships from large quantities of data through computational methods. With the limited ability of traditional association-based computational methods to discover causal relationships, CSD methodologies are gaining popularity. The goal of the study was to systematically examine whether (i) CSD methods can discover the known causal relationships from observational clinical data and (ii) to offer guidance to accurately discover known causal relationships. We used Alzheimer's disease (AD), a complex progressive disease, as a model because the well-established evidence provides a "gold-standard" causal graph for evaluation. We evaluated two CSD methods, Fast Causal Inference (FCI) and Fast Greedy Equivalence Search (FGES) in their ability to discover this structure from data collected by the Alzheimer's Disease Neuroimaging Initiative (ADNI). We used structural equation models (which is not designed for CSD) as control. We applied these methods under three scenarios defined by increasing amounts of background knowledge provided to the methods. The methods were evaluated by comparing the resulting causal relationships with the "gold standard" graph that was constructed from literature. Dedicated CSD methods managed to discover graphs that nearly coincided with the gold standard. For best results, CSD algorithms should be used with longitudinal data providing as much prior knowledge as possible.

Algorithms

Aged

Aged, 80 and over

Alzheimer Disease - diagnosis

Alzheimer Disease - etiology

Alzheimer Disease - pathology

Amyloid beta-Peptides - analysis

Apolipoprotein E4 - genetics

Apolipoproteins E - genetics

Biomarkers - analysis

Brain - diagnostic imaging

Brain - pathology

Data Interpretation, Statistical

Datasets as Topic

Female

Humans

Latent Class Analysis

Longitudinal Studies

Magnetic Resonance Imaging - statistics & numerical data

Male

Models, Neurological

Neuroimaging - statistics & numerical data

Observational Studies as Topic

Positron-Emission Tomography - statistics & numerical data

tau Proteins - analysis

Details

Title: Subtitle: Challenges and Opportunities with Causal Discovery Algorithms: Application to Alzheimer's Pathophysiology
Creators: Xinpeng Shen - University of Minnesota
Sisi Ma - University of Minnesota
Prashanthi Vemuri - Mayo Clinic, Rochester, MN 55905 (USA)
Gyorgy Simon - University of Minnesota
Alzheimer’s Disease Neuroimaging Initiative
Contributors: Hristina Koleva (Contributor) - University of Iowa, Psychiatry
Resource Type: Journal article
Publication Details: Scientific reports, Vol.10(1), pp.2975-2975
DOI: 10.1038/s41598-020-59669-x
PMID: 32076020
PMCID: PMC7031278
NLM abbreviation: Sci Rep
ISSN: 2045-2322
eISSN: 2045-2322
Grant note: UL1 TR002369 / NCATS NIH HHS R01 NS097495 / NINDS NIH HHS P30 AG066462 / NIA NIH HHS P30 AG062715 / NIA NIH HHS U01 AG024904 / NIA NIH HHS R01 AG056366 / NIA NIH HHS
Language: English
Date published: 02/19/2020
Academic Unit: Psychiatry
Record Identifier: 9984293653702771

Metrics

28 Record Views

116 Times Cited - Web of Science