A causal inference and Bayesian optimisation framework for modelling multi-trait relationships—Proof-of-concept using Brassica napus seed yield under controlled conditions

A - Papers appearing in refereed journals

Calderwood, A., Siles-Suarez, L., Eastmond, P. J., Kurup, S. and Morris, R. 2023. A causal inference and Bayesian optimisation framework for modelling multi-trait relationships—Proof-of-concept using Brassica napus seed yield under controlled conditions. PLOS ONE. 18 (9), p. 0290429. https://doi.org/10.1371/journal.pone.0290429

AuthorsCalderwood, A., Siles-Suarez, L., Eastmond, P. J., Kurup, S. and Morris, R.
Abstract

The improvement of crop yield is a major breeding target and there is a long history of research that has focussed on unravelling the mechanisms and processes that contribute to yield. Quantitative prediction of the interplay between morphological traits, and the effects of these trait-trait relationships on seed production remains, however, a challenge. Consequently, the extent to which crop varieties optimise their morphology for a given environment is largely unknown. This work presents a new combination of existing methodologies by framing crop breeding as an optimisation problem and evaluates the extent to which existing varieties exhibit optimal morphologies under the test conditions. In this proof-of-concept study using spring and winter oilseed rape plants grown under greenhouse conditions, we employ causal inference to model the hierarchically structured effects of 27 morphological yield traits on each other. We perform Bayesian optimisation of seed yield, to identify and quantify the morphologies of ideotype plants, which are expected to be higher yielding than the varieties in the studied panels. Under the tested growth conditions, we find that existing spring varieties occupy the optimal regions of trait-space, but that potentially high yielding strategies are unexplored in extant winter varieties. The same approach can be used to evaluate trait (morphology) space for any environment.

Year of Publication2023
JournalPLOS ONE
Journal citation18 (9), p. 0290429
Digital Object Identifier (DOI)https://doi.org/10.1371/journal.pone.0290429
Open accessPublished as ‘gold’ (paid) open access
FunderBiotechnology and Biological Sciences Research Council
Funder project or codeTailoring Plant Metabolism (TPM) - Work package 1 (WP1) - High value lipids for health and industry
Brassica Rapeseed And Vegetable Optimisation (BRAVO)
Publisher's version
Output statusPublished
Publication dates
Online01 Sep 2023
Publication process dates
Accepted09 Aug 2023
PublisherPublic Library of Science (PLOS)
ISSN1932-6203

Permalink - https://repository.rothamsted.ac.uk/item/98y6v/a-causal-inference-and-bayesian-optimisation-framework-for-modelling-multi-trait-relationships-proof-of-concept-using-brassica-napus-seed-yield-under-controlled-conditions

84 total views
19 total downloads
6 views this month
0 downloads this month
Download files as zip