Combination therapy has gained popularity in cancer treatment as it enhances the treatment efficacy and overcomes drug resistance. Although machine learning (ML) techniques have become an indispensable tool for discovering new drug combinations, the data on drug combination therapy currently available may be insufficient to build high-precision models. We developed a data augmentation protocol to unbiasedly scale up the existing anti-cancer drug synergy dataset. Using a new drug similarity metric, we augmented the synergy data by substituting a compound in a drug combination instance with another molecule that exhibits highly similar pharmacological effects. Using this protocol, we were able to upscale the AZ-DREAM Challenges dataset from 8798 to 6,016,697 drug combinations. Comprehensive performance evaluations show that ML models trained on the augmented data consistently achieve higher accuracy than those trained solely on the original dataset. Our data augmentation protocol provides a systematic and unbiased approach to generating more diverse and larger-scale drug combination datasets, enabling the development of more precise and effective ML models. The protocol presented in this study could serve as a foundation for future research aimed at discovering novel and effective drug combinations for cancer treatment.
Combination therapy is a promising strategy for confronting the complexity of cancer. However, experimental exploration of the vast space of potential drug combinations is costly and unfeasible. Therefore, computational methods for predicting drug synergy are much needed for narrowing down this space, especially when examining new cellular contexts. Here, we thus introduce CCSynergy, a flexible, context aware and integrative deep-learning framework that we have established to unleash the potential of the Chemical Checker extended drug bioactivity profiles for the purpose of drug synergy prediction. We have shown that CCSynergy enables predictions of superior accuracy, remarkable robustness and improved context generalizability as compared to the state-of-the-art methods in the field. Having established the potential of CCSynergy for generating experimentally validated predictions, we next exhaustively explored the untested drug combination space. This resulted in a compendium of potentially synergistic drug combinations on hundreds of cancer cell lines, which can guide future experimental screens.
more » « less- NSF-PAR ID:
- 10387391
- Publisher / Repository:
- Oxford University Press
- Date Published:
- Journal Name:
- Briefings in Bioinformatics
- Volume:
- 24
- Issue:
- 1
- ISSN:
- 1467-5463
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract -
more » « less
Effective cancer therapies often demand delivery of combinations of drugs to inhibit multidrug resistance through synergism, and the development of multifunctional nanovehicles with enhanced drug loading and delivery efficiency for combination therapy is currently a major challenge in nanotechnology. However, such combinations are more challenging to administer than single drugs and can require multipronged approaches to delivery. In addition to being stable and biodegradable, vehicles for such therapies must be compatible with both hydrophobic and hydrophilic drugs, and release drugs at sustained therapeutic levels. Here, we report synthesis of porous silicon nanoparticles conjugated with gold nanorods [composite nanoparticles (cNPs)] and encapsulate them within a hybrid polymersome using double-emulsion templates on a microfluidic chip to create a versatile nanovehicle. This nanovehicle has high loading capacities for both hydrophobic and hydrophilic drugs, and improves drug delivery efficiency by accumulating at the tumor after i.v. injection in mice. Importantly, a triple-drug combination suppresses breast tumors by 94% and 87% at total dosages of 5 and 2.5 mg/kg, respectively, through synergy. Moreover, the cNPs retain their photothermal properties, which can be used to significantly inhibit multidrug resistance upon near-infrared laser irradiation. Overall, this work shows that our nanovehicle has great potential as a drug codelivery nanoplatform for effective combination therapy that is adaptable to other cancer types and to molecular targets associated with disease progression.
-
more » « less
Abstract Designing multi‐drug regimens often involves target‐ and synergy prediction‐based drug selection, and subsequent dose escalation to achieve the maximum tolerated dose (MTD) of each drug. This approach may improve efficacy, but not the optimal efficacy and often substantially increases toxicity. Drug interactions depend on many pathways in the omics networks and further complicate the design process. The virtually infinite drug–dose parameter space cannot be reconciled using conventional approaches, which are largely based on prediction. This barrier at least partially accounts for the low response rates that are observed with conventional mono‐ and combinatorial chemotherapy. A combination of nonstandard therapies for colorectal cancer (AGCH: adriamycin, gemcitabine, cisplatin, and herceptin) at ¼ MTD is used to treat the rats, the tumor response rates varied in a wide range. Some of the tumor response rates are close to that of control group. This work harnesses an artificial intelligence (AI) platform that is mechanism free and can dynamically optimize combinatorial therapy in rats. The individually optimized AGCH regimen reveal starkly different drug–dose parameters, which can converge each rat toward the same and low tumor response rate. Importantly, this AI‐based drug–dose optimization technology is an actionable platform, which can achieve N‐of‐1 therapy.
-
more » « less
Abstract Vascular endothelial growth factor (VEGF) is a strong promoter of angiogenesis in tumors, and anti‐VEGF treatment, such as a humanized antibody to VEGF, is clinically used as a monotherapy or in combination with chemotherapy to treat cancer patients. However, this approach is not effective in all patients or cancer types. To better understand the heterogeneous responses to anti‐VEGF and the synergy between anti‐VEGF and other anticancer therapies, we constructed a computational model characterizing angiogenesis‐mediated growth of
in vivo mouse tumor xenografts. The model captures VEGF‐mediated cross‐talk between tumor cells and endothelial cells and is able to predict the details of molecular‐ and cellular‐level dynamics. The model predictions of tumor growth in response to anti‐VEGF closely match the quantitative measurements from multiple preclinical mouse studies. We applied the model to investigate the effects of VEGF‐targeted treatment on tumor cells and endothelial cells. We identified that tumors with lower tumor cell growth rate and higher carrying capacity have a stronger response to anti‐VEGF treatment. The predictions indicate that the variation of tumor cell growth rate can be a main reason for the experimentally observed heterogeneous response to anti‐VEGF. In addition, our simulation results suggest a new synergy mechanism where anticancer therapy can enhance anti‐VEGF simply through reducing the tumor cell growth rate. Overall, this work generates novel insights into the heterogeneous response to anti‐VEGF treatment and the synergy of anti‐VEGF with other therapies, providing a tool that be further used to test and optimize anticancer therapy. -
more » « less
Abstract Motivation Finding driver genes that are responsible for the aberrant proliferation rate of cancer cells is informative for both cancer research and the development of targeted drugs. The established experimental and computational methods are labor-intensive. To make algorithms feasible in real clinical settings, methods that can predict driver genes using less experimental data are urgently needed.
Results We designed an effective feature selection method and used Support Vector Machines (SVM) to predict the essentiality of the potential driver genes in cancer cell lines with only 10 genes as features. The accuracy of our predictions was the highest in the Broad-DREAM Gene Essentiality Prediction Challenge. We also found a set of genes whose essentiality could be predicted much more accurately than others, which we called Accurately Predicted (AP) genes. Our method can serve as a new way of assessing the essentiality of genes in cancer cells.
Availability and implementation The raw data that support the findings of this study are available at Synapse. https://www.synapse.org/#! Synapse: syn2384331/wiki/62825. Source code is available at GitHub. https://github.com/GuanLab/DREAM-Gene-Essentiality-Challenge.
Supplementary information Supplementary data are available at Bioinformatics online.
