AITL: Adversarial Inductive Transfer Learning with input and output space adaptation for pharmacogenomics

Abstract

Motivation The goal of pharmacogenomics is to predict drug response in patients using their single- or multi-omics data. A major challenge is that clinical data (i.e. patients) with drug response outcome is very limited, creating a need for transfer learning to bridge the gap between large pre-clinical pharmacogenomics datasets (e.g. cancer cell lines), as a source domain, and clinical datasets as a target domain. Two major discrepancies exist between pre-clinical and clinical datasets: (i) in the input space, the gene expression data due to difference in the basic biology, and (ii) in the output space, the different measures of the drug response. Therefore, training a computational model on cell lines and testing it on patients violates the i.i.d assumption that train and test data are from the same distribution. Results We propose Adversarial Inductive Transfer Learning (AITL), a deep neural network method for addressing discrepancies in input and output space between the pre-clinical and clinical datasets. AITL takes gene expression of patients and cell lines as the input, employs adversarial domain adaptation and multi-task learning to address these discrepancies, and predicts the drug response as the output. To the best of our knowledge, AITL is the first adversarial inductive transfer learning method to address both input and output discrepancies. Experimental results indicate that AITL outperforms state-of-the-art pharmacogenomics and transfer learning baselines and may guide precision oncology more accurately. Availability and implementation https://github.com/hosseinshn/AITL. Supplementary information Supplementary data are available at Bioinformatics online.

Super.FELT: supervised feature extraction learning using triplet loss for drug response prediction with multi-omics data

“…Recent studies have proposed models with adversarial networks by training both in vitro and in vivo datasets and obtained increased performance compared with that of models trained using only in vitro datasets [51]. This approach is helpful when appropriate in vivo datasets for a given drug are available.…”

Transfer Learning of Clinical Outcomes with Molecular Data, Principles and Perspectives

“…As the authors of AITL (see below) point out, considering in-vitro to in-vivo translation: "Two major discrepancies exist between preclinical and clinical datasets: (i) in the input space, the gene expression data, due to difference in the basic biology, and (ii) in the output space, the different measures of the drug response. Therefore, training a computational model on cell lines and testing it on patients violates the i.i.d assumption that train and test data are from the same distribution" (Sharifi-Noghabi et al, 2020). Other frequently encountered gaps hindering transfer are the species gap and the complex relationships between the molecular entities that may be measured (see above).…”

“…By default, we follow Sharifi-Noghabi et al (2020) in adopting the terminology of Pan and Yang (2010), which is a widely cited review and reference in the field of transfer learning. In their paper, transfer learning is defined in terms of source and target domains (of features with probability distributions associated with these), as well as source and target tasks (mapping features to labels using predictors) so that the predictor in the target domain uses some knowledge from the source domain.…”

TUGDA: task uncertainty guided domain adaptation for robust generalization of cancer drug response prediction from in vitro to in vivo settings

“…Previous works ( Geeleher et al , 2014 , 2017 ; Sakellaropoulos et al , 2019 ) assumed that batch effects were the main origin of differences to correct for between models, without directly addressing biological variations. Recently, some methods have sought to use domain adaptation (DA) techniques to bridge the in vitro to in vivo gap ( Mourragui et al , 2019 , 2020 ; Sharifi-Noghabi et al , 2020 ).…”

“…A refinement of this idea, TRANSACT ( Mourragui et al , 2020 ), uses Kernel-PCA based sub-space alignment to further capture non-linear relationships between samples from in vitro and in vivo domains. However, to learn the similarity between domains, existing DA methods either do not take into account the conditional distributions ( and for drug response Y given gene expression X in source s and target t ), obtaining a subset of shared features that might be unrelated to drug response ( Mourragui et al , 2019 , 2020 ), or rely on the covariate-shift assumption ( Sharifi-Noghabi et al , 2020 ), where marginal distributions for features ( and , for tasks/domains s and t ) are allowed to vary while the conditional distribution for drug response is assumed to be the same ( ) ( Kouw and Loog, 2021 ; Zhao et al , 2019 ). This assumption can often lead to NT ( Rampášek, 2020 ; Zhao et al , 2019 ) when e.g.…”