On board with COMET to improve omics prediction models
Related Articles
A machine learning approach to leveraging electronic health records for enhanced omics analysis
Omics studies produce a large number of measurements, enabling the development, validation and interpretation of systems-level biological models. Large cohorts are required to power these complex models; yet, the cohort size remains limited due to clinical and budgetary constraints. We introduce clinical and omics multimodal analysis enhanced with transfer learning (COMET), a machine learning framework that incorporates large, observational electronic health record databases and transfer learning to improve the analysis of small datasets from omics studies. By pretraining on electronic health record data and adaptively blending both early and late fusion strategies, COMET overcomes the limitations of existing multimodal machine learning methods. Using two independent datasets, we showed that COMET improved the predictive modelling performance and biological discovery compared with the analysis of omics data with traditional methods. By incorporating electronic health record data into omics analyses, COMET enables more precise patient classifications, beyond the simplistic binary reduction to cases and controls. This framework can be broadly applied to the analysis of multimodal omics studies and reveals more powerful biological insights from limited cohort sizes.
Targeting Chk1 and Wee1 kinases enhances radiosensitivity of 2D and 3D head and neck cancer models to X-rays and low/high-LET protons
Ionising radiation causes the introduction of DNA damage, more specifically double strand breaks (DSBs) and complex DNA damage (CDD), that induces cancer cell death leading to the therapeutic effect. To combat this, cells activate arrest at the G2/M checkpoint to allow for effective DNA damage repair, coordinated by the Chk1 and Wee1 protein kinases. Therefore, Chk1 and Wee1 are considered promising therapeutic targets to enhance the effectiveness of radiotherapy in cancer cell killing. Here, we have analysed the response of head and neck squamous cell carcinoma (HNSCC) cell lines, spheroids and patient-derived organoids to X-rays and proton beam therapy (PBT) in the presence of either a Chk1 (MK-8776) or a Wee1 (MK-1775) inhibitor. We demonstrate that inhibitors of Chk1 or Wee1 can significantly enhance the radiosensitivity of both 2D and 3D models of HNSCC to X-rays and PBT (performed at both low and high ionisation densities), and that this effect is caused through abrogation of the G2/M checkpoint causing the persistence of DSBs. Our results therefore suggest that targeting Chk1 and Wee1 kinases in combination with X-rays and PBT could represent a promising therapeutic avenue to enhance the clinical efficacy of HNSCC treatment.
AI can outperform humans in predicting correlations between personality items
We assess the abilities of both specialized deep neural networks, such as PersonalityMap, and general LLMs, including GPT-4o and Claude 3 Opus, in understanding human personality by predicting correlations between personality questionnaire items. All AI models outperform the vast majority of laypeople and academic experts. However, we can improve the accuracy of individual correlation predictions by taking the median prediction per group to produce a “wisdom of the crowds” estimate. Thus, we also compare the median predictions from laypeople, academic experts, GPT-4o/Claude 3 Opus, and PersonalityMap. Based on medians, PersonalityMap and academic experts surpass both LLMs and laypeople on most measures. These results suggest that while advanced LLMs make superior predictions compared to most individual humans, specialized models like PersonalityMap can match even expert group-level performance in domain-specific tasks. This underscores the capabilities of large language models while emphasizing the continued relevance of specialized systems as well as human experts for personality research.
A data-driven generative strategy to avoid reward hacking in multi-objective molecular design
Molecular design using data-driven generative models has emerged as a promising technology, impacting various fields such as drug discovery and the development of functional materials. However, this approach is often susceptible to optimization failure due to reward hacking, where prediction models fail to extrapolate, i.e., fail to accurately predict properties for designed molecules that considerably deviate from the training data. While methods for estimating prediction reliability, such as the applicability domain (AD), have been used for mitigating reward hacking, multi-objective optimization makes it challenging. The difficulty arises from the need to determine in advance whether the multiple ADs with some reliability levels overlap in chemical space, and to appropriately adjust the reliability levels for each property prediction. Herein, we propose a reliable design framework to perform multi-objective optimization using generative models while preventing reward hacking. To demonstrate the effectiveness of the proposed framework, we designed candidates for anticancer drugs as a typical example of multi-objective optimization. We successfully designed molecules with high predicted values and reliabilities, including an approved drug. In addition, the reliability levels can be automatically adjusted according to the property prioritization specified by the user without any detailed settings.
A unified deep framework for peptide–major histocompatibility complex–T cell receptor binding prediction
Antigen peptides that are presented by a major histocompatibility complex (MHC) and recognized by a T cell receptor (TCR) have an essential role in immunotherapy. Although substantial progress has been made in predicting MHC presentation, accurately predicting the binding interactions between antigen peptides, MHCs and TCRs remains a major computational challenge. In this paper, we propose a unified deep framework (called UniPMT) for peptide, MHC and TCR binding prediction to predict the binding between the peptide and the CDR3 of TCR β in general, presented by class I MHCs. UniPMT is comprehensively validated by a series of experiments and achieved state-of-the-art performance in the peptide–MHC–TCR, peptide–MHC and peptide–TCR binding prediction tasks with up to 15% improvements in area under the precision–recall curve taking the peptide–MHC–TCR binding prediction task as an example. In practical applications, UniPMT shows strong predictive power, correlates well with T cell clonal expansion and outperforms existing methods in neoantigen-specific binding prediction with up to 17.62% improvements in area under the precision–recall curve on experimentally validated datasets. Moreover, UniPMT provides interpretable insights into the identification of key binding sites and the quantification of peptide–MHC–TCR binding probabilities. In summary, UniPMT shows great potential to serve as a useful tool for antigen peptide discovery, disease immunotherapy and neoantigen vaccine design.
Responses