Data-Driven Modelling of Gene Expression States in Breast Cancer and their Prediction from Routine Whole Slide Images

Dawood, Muhammad; Eastwood, Mark; Jahanifar, Mostafa; Young, Lawrence S.; Ben‐Hur, Asa; Branson, Kristin; Jones, Louise; Rajpoot, Nasir M.; Minhas, Fayyaz ul Amir Afsar

doi:10.1101/2023.04.14.536756

Cited by 3 publications

(7 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this study, we proposed a novel approach to predict the sensitivity of breast cancer tumours to various drugs from histological patterns in their whole slide images (WSI). We employed our 𝑆𝑙𝑖𝑑𝑒𝐺𝑟𝑎𝑝ℎ pipeline [39] to explicitly model the local and global histological patterns in the WSI. More specially, we construct a graph representation of the WSI first and then train a graph neural network (GNN) that not only predicts WSI-level sensitivity but also highlights spatially resolved contribution of different WSI regions towards the predicted sensitivity estimate.…”

Section: Discussionmentioning

confidence: 99%

“…To explore the association between cellular and histological patterns contained in WSIs and patient tumour's sensitivity to different drugs, we propose an end-to-end DL pipeline that takes WSI of a patient as input and predicts the sensitivity of 427 compounds as output. An overview of the proposed framework is provided in Fig 1C . We employed our in-house 𝑆𝑙𝑖𝑑𝑒𝐺𝑟𝑎𝑝ℎ pipeline [39] that first constructs a graph representation of the WSI and then uses a graph neural network (GNN) to predict node-level (patch-level) and WSI-level sensitivity of a patient to all the compounds (compounds listed in Table S1). The node-level scores are then used to identify regions within the WSI that contribute to high or low sensitivity.…”

Section: Analytical Pipeline For Whole Slide Image Analysis and Predi...mentioning

confidence: 99%

“…We utilize a Graph Neural Network (GNN) to predict node-level and WSI-level sensitivity of patients to a set of drugs 𝐷 from their corresponding WSI-graphs. More specifically, we used 𝑆𝑙𝑖𝑑𝑒𝐺𝑟𝑎𝑝ℎ with slightly modified architecture [39]. The multi-output GNN model predicts patch-level and WSI-level sensitivity of patients to 𝐷 different drugs in an end-to-end manner.…”

Section: Prediction Of Drug Sensitivity Using Graph Neural Network (Gnn)mentioning

confidence: 99%

“…Moreover, histology image based prediction of mutation and expression profile of different genes [33]- [36] and prediction of molecular markers and pathways [37], [38], has been achieved using deep learning. Recently, a deep learning model has been proposed for predicting breast cancer patients gene expression state from WSIs [39].…”

Section: Introductionmentioning

confidence: 99%

“…In this work, we investigate the association between cellular and morphometric patterns contained in the digitised WSIs of routine H&E tissue slides of breast cancer tumours and their drug sensitivity profiles. We employ the method proposed in [39] for predicting patient's likelihood of response to treatment to specific drugs. The framework proposed in this study (see Fig 1) offers several possible advantages.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Cancer drug sensitivity prediction from routine histology images

Dawood,

Vu,

Young

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Drug sensitivity prediction models can aid in personalising cancer therapy, biomarker discovery, and drug design. Such models require survival data from randomized controlled trials which can be time consuming and expensive. In this proof-of-concept study, we demonstrate for the first time that deep learning can link histological patterns in whole slide images (WSIs) of Haematoxylin & Eosin (H&E) stained breast cancer sections with drug sensitivities inferred from cell lines. We employ patient-wise drug sensitivities imputed from gene expression based mapping of drug effects on cancer cell lines to train a deep learning model that predicts sensitivity to multiple drugs from WSIs. We show that it is possible to use routine WSIs to predict the drug sensitivity profile of a cancer patient for a number of approved and experimental drugs. We also show that the proposed approach can identify cellular and histological patterns associated with drug sensitivity profiles of cancer patients.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Analytical Pipeline For Whole Slide Image Analysis and Predi...mentioning

confidence: 99%

Section: Prediction Of Drug Sensitivity Using Graph Neural Network (Gnn)mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Cancer drug sensitivity prediction from routine histology images

Dawood,

Vu,

Young

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Cancer drug sensitivity prediction from routine histology images

Dawood

Young

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Buyer Beware: confounding factors and biases abound when predicting omics-based biomarkers from histological images

Dawood,

Branson,

Tejpar

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

SummaryBackgroundRecent advancements in computational pathology have introduced deep learning methods to predict genomic, transcriptomic and molecular biomarkers from routine histology whole slide images (WSIs) for cancer diagnosis, prognosis, and treatment. However, existing methods often overlook the critical role of co-dependencies among biomarker statuses during training and inference. We hypothesize that this oversight results in models that predict the combined effect of multiple interdependent biomarkers rather than individual statuses independently, akin to attributing the quality of an orchestral symphony to a single instrument, highlighting limitations of current predictors.MethodsUsing large datasets (n = 8,221 patients), we conducted statistical co-dependence testing to demonstrate significant interdependencies among biomarker statuses in training datasets. Following standard protocols, we trained two machine learning models to predict biomarkers from WSIs achieving or matching state-of-the-art predictive performance. We then employed permutation testing and stratification analysis to evaluate their predictive quality based on the principle of conditional independence, i.e., if a model accurately captures the phenotypic influence of a specific biomarker independent of other biomarkers, its performance should remain consistent across subgroups of patients stratified by other biomarkers, aligning with its overall performance on the entire dataset.FindingsOur statistical analysis reveals significant interdependencies among biomarkers, reflecting expected co-occurrence and mutual exclusivity patterns influenced by pathological and biological processes that are consistent across datasets, as well as sampling artefacts that can be different across datasets. Our results indicate that the predictive quality of an image-based predictor for a biomarker is contingent on the status of other biomarkers, revealing that models capture aggregated influences rather than predicting individual statuses independently. For example, mutation predictions are confounded by the overall tumour mutation burden. We also show that, due to the presence of such correlations, deep learning models may not offer significant advantages in predicting certain biomarkers in comparison to simply using pathologist-assigned grades for their prediction.InterpretationWe show that current deep learning models in computational pathology fall short in isolating individual biomarker effects, leading to confounded and less precise predictions. Our findings suggest revisiting model training protocols to recognize and adjust for biomarker interdependencies at all development stages—from problem definition to usage guidelines. This involves selecting diverse datasets to reflect clinical heterogeneity, defining prediction variables or grouping patients based on co-dependencies, designing models to disentangle complex relationships, and stringent stratification testing. Clinically, failure to account for interdependencies may lead to suboptimal decisions, necessitating appropriate usage guidelines for predictive models.

show abstract

Data-Driven Modelling of Gene Expression States in Breast Cancer and their Prediction from Routine Whole Slide Images

Cited by 3 publications

References 53 publications

Cancer drug sensitivity prediction from routine histology images

Cancer drug sensitivity prediction from routine histology images

Cancer drug sensitivity prediction from routine histology images

Buyer Beware: confounding factors and biases abound when predicting omics-based biomarkers from histological images

Contact Info

Product

Resources

About