Re-estimation-verified selection of small parameter subsets for systems biology models

Abstract Mathematical models of biological systems have far more unknown parameters than experiments can determine, and estimating all of them is slow, unstable and yields values that the data cannot pin down. A common remedy is to estimate only a few and keep the rest at published values, yet the methods that choose those few do not check that the simplified model still behaves like the original one. This work presents a selection method that accepts a small set of parameters only after verifying that re-estimating them reproduces the response of the full model to random changes of all its parameters, and then discards every parameter that is not needed. Applied to 34 published models with real data, from cell signalling to epidemics, it found that two parameters usually suffice, and that the data determine them. Against four established criteria, its sets were sufficient and determined by the data in at least as many models. The mathematical results behind the method are verified by a computer proof assistant. Recalibrating a published model with new data can thus start from a few checked parameters instead of a median of fifteen.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-17
DOI
https://doi.org/10.21203/rs.3.rs-11750537/v1
Primary Topic
Gene Regulatory Network Analysis
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Re-estimation-verified selection of small parameter subsets for systems biology models

Cesar Antonio Martin Moreno, Monica Maria Miranda Ramos, Adriana Allison Aguirre Alonso, Carlos Alberto Salazar López
Research Square
Gene Regulatory Network Analysis
preprint

Re-estimation-verified selection of small parameter subsets for systems biology models

Cesar Antonio Martin Moreno, Monica Maria Miranda Ramos, Adriana Allison Aguirre Alonso, Carlos Alberto Salazar López
preprint en

Abstract

Abstract Mathematical models of biological systems have far more unknown parameters than experiments can determine, and estimating all of them is slow, unstable and yields values that the data cannot pin down. A common remedy is to estimate only a few and keep the rest at published values, yet the methods that choose those few do not check that the simplified model still behaves like the original one. This work presents a selection method that accepts a small set of parameters only after verifying that re-estimating them reproduces the response of the full model to random changes of all its parameters, and then discards every parameter that is not needed. Applied to 34 published models with real data, from cell signalling to epidemics, it found that two parameters usually suffice, and that the data determine them. Against four established criteria, its sets were sufficient and determined by the data in at least as many models. The mathematical results behind the method are verified by a computer proof assistant. Recalibrating a published model with new data can thus start from a few checked parameters instead of a median of fifteen.

Research Square
Escuela Superior Politecnica del Litoral (EC)
Gene Regulatory Network Analysis
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon