Differential Private Deep Learning Models for Analyzing Breast Cancer Omics Data

Abstract

Proper analysis of high-dimensional human genomic data is necessary to increase human knowledge about fundamental biological questions such as disease associations and drug sensitivity. However, such data contain sensitive private information about individuals and can be used to identify an individual (i.e., privacy violation) uniquely. Therefore, raw genomic datasets cannot be publicly published or shared with researchers. The recent success of deep learning (DL) in diverse problems proved its suitability for analyzing the high volume of high-dimensional genomic data. Still, DL-based models leak information about the training samples. To overcome this challenge, we can incorporate differential privacy mechanisms into the DL analysis framework as differential privacy can protect individuals’ privacy. We proposed a differential privacy based DL framework to solve two biological problems: breast cancer status (BCS) and cancer type (CT) classification, and drug sensitivity prediction. To predict BCS and CT using genomic data, we built a differential private (DP) deep autoencoder (dpAE) using private gene expression datasets that performs low-dimensional data representation learning. We used dpAE features to build multiple DP binary classifiers to predict BCS and CT in any individual. To predict drug sensitivity, we used the Genomics of Drug Sensitivity in Cancer (GDSC) dataset. We extracted GDSC’s dpAE features to build our DP drug sensitivity prediction model for 265 drugs. Evaluation of our proposed DP framework shows that it achieves improved prediction performance in predicting BCS, CT, and drug sensitivity than the previously published DP work.

Risk Stratification for Breast Cancer Patient by Simultaneous Learning of Molecular Subtype and Survival Outcome Using Genetic Algorithm-Based Gene Set Selection

“…Although the rank-based methods are more robust to outliers, platforms and batches, there may be a loss of quantity information by converting the expression quantity to a rank. Therefore, a method using the expression quantity of the gene itself or a method applying the differential privacy mechanism for individuals privacy [ 79 ] could be more effective for representing individual patients. Next, interactions among genes were not directly considered.…”

Privacy-Preserving Deep Learning Model for Covid-19 Disease Detection

“…The healthcare applications like cancer detection, diabetic retinopathy grading analysis, etc. (Sai Venkatesh et al, 2022;Chilukoti et al, 2022;Islam et al, 2022) can use the proposed privacy-preserving deep learning model with a DP Adam optimizer to achieve better performance. Furthermore, the proposed work can also be applied in federated learning settings by making individual privacy-preserving models before aggregating them into the central server (Yang et al, 2019;Wei et al, 2020).…”

PPML-Omics: A privacy-preserving federated machine learning method protects patients’ privacy in omic data

“…Apart from that, only swarm learning (32) discussed the application of FL with blockchain technology in the analysis of omic data. But with the notation of DP, to our knowledge, only Islam et al (50) discussed the application of DP in DL for breast cancer status and cancer type classification, and drug sensitivity prediction, and only one work discussed the application of FL-DP in cancer prediction as a solution to the competition hosted by iDASH (integrating Data for Analysis, Anonymization, SHaring) National Center for Biomedical Computing in 2020. Apart from that, there is no more work that has systematically studied and delved into the privacy protection of sequencing data from a bigger picture with the DP notation, even though raw sequencing data contains much more private information about patients than SNPs and GWAS.…”