Secure multiparty computation (MPC) enables privacy-preserving data analysis across distributed computing parties, but its practicality remains limited by the cost and inaccuracy of evaluating complex nonlinear functions. Although secure multiplications are efficiently supported by Beaver multiplication triples-a cornerstone technique in MPC-existing protocols rely on polynomial approximations to evaluate more sophisticated functions, often incurring significant computational overhead and numerical imprecision. We present Decor, a framework that generalizes the core principle behind the Beaver triples to a broad class of nonlinear functions, enabling the construction of efficient MPC primitives. Decor delegates costly nonlinear operations to computations involving only random values, which can be executed in an offline preprocessing phase by a trusted dealer without access to private data. This design enables efficient and accurate evaluation of diverse functions, including trigonometric, hyperbolic, exponential, and sigmoid functions, and introduces a new general-purpose function approximation method for MPC based on Fourier series. Our experiments show that Decor achieves orders-ofmagnitude improvements in accuracy while maintaining comparable or faster runtimes than existing approaches, leading to substantial utility gains for downstream applications such as implicit neural representation of images and logistic model estimation in genome-wide association studies. By demonstrating how randomized preprocessing can yield enhanced MPC primitives, Decor establishes a new framework for practical, privacy-preserving computation.

Scattered between many healthcare providers across the US, Electronic Health Records (EHR) are extensively used for research purposes. Collaboration and sharing of EHRs between multiple institutions often provide access to more diverse datasets and a chance to conduct comprehensive studies. However, these collaboration efforts are usually hindered by privacy issues that render the pooling of such data at a centralized database impossible. Furthermore, EHRs are often incomplete and require statistical imputation prior to the study. To enable collaborative studies on top of incomplete, private EHRs, here we provide a provably secure solution built with secure multiparty computation (SMC) that provides practical runtimes and accuracy on par with the state-of-the-art, non-secure equivalents. Our solution enables the utilization of distributed datasets as a whole to impute the missing data and conduct collective studies between non-trusting private data proprietors. We demonstrate its effectiveness on various synthetic and real-world datasets, and show that our solution can significantly improve the classification of high-risk patient outcomes during ICU admission.

We present Shechi, an easy-to-use programming framework for secure high-performance computing on distributed datasets. Shechi automatically converts Pythonic code into a secure distributed equivalent using multiparty homomorphic encryption (MHE), combining homomorphic encryption (HE) and secure multiparty computation (SMC) techniques to enable efficient distributed computation. Shechi abstracts away considerations about the private and distributed aspects of the input data from end users through a familiar Pythonic syntax. Our framework introduces new data types for the efficient handling of distributed data as well as systematic compiler optimizations for cryptographic and distributed computations. We evaluate Shechi on a wide range of applications, including principal component analysis and complex genomic analysis tasks. Our results demonstrate Shechi’s ability to uncover optimizations missed even by expert developers, achieving up to 15× runtime improvements over the prior state-of-the-art solutions and a 40-fold improvement in code expressiveness compared to code manually optimized by experts. Shechi represents the first MHE compiler, extending secure computation frameworks to the analysis of sensitive distributed datasets.

Dynamic programming (DP) is a fundamental algorithmic strategy that decomposes large problems into manageable subproblems. It is a cornerstone of many important computational methods in diverse fields, especially in the field of computational genomics, where it is used for sequence comparison. However, as the scale of the data keeps increasing, these algorithms are becoming a major computational bottleneck, and there is a need for strategies that can improve their performance. Here, we present Vectron, a novel auto-vectorization suite that targets array-based DP implementations written in Python and converts them to efficient vectorized counterparts that can efficiently process multiple problem instances in parallel. Leveraging Single Instruction Multiple Data (SIMD) capabilities in modern CPUs, along with Graphics Processing Units (GPUs), Vectron delivers significant speedups, ranging from 10% to more than 20x, over the conventional C++ implementations and manually vectorized and domain-specific state-of-the-art implementations, without necessitating large algorithm or code changes. Vectron's generality enables automatic vectorization of any array-based DP algorithm and, as a result, presents an attractive solution to optimization challenges inherent to DP algorithms.

Domain-specific languages (DSLs) are able to provide intuitive high-level abstractions that are easy to work with while attaining better performance than general-purpose languages. Yet, implementing new DSLs is a burdensome task. As a result, new DSLs are usually embedded in general-purpose languages. While low-level languages like C or C++ often provide better performance as a host than high-level languages like Python, high-level languages are becoming more prevalent in many domains due to their ease and flexibility. Here, we present Codon, a domain-extensible compiler and DSL framework for high-performance DSLs with Python's syntax and semantics. Codon builds on previous work on ahead-of-time type checking and compilation of Python programs and leverages a novel intermediate representation to easily incorporate domain-specific optimizations and analyses. We showcase and evaluate several compiler extensions and DSLs for Codon targeting various domains, including bioinformatics, secure multi-party computation, block-based data compression and parallel programming, showing that Codon DSLs can provide benefits of familiar high-level languages and achieve performance typically only seen with low-level languages, thus bridging the gap between performance and usability.

Secure multiparty computation (MPC) is a cryptographic tool that allows computation on top of sensitive biomedical data without revealing private information to the involved entities. Here, we introduce Sequre, an easy-to-use, high-performance framework for developing performant MPC applications. Sequre offers a set of automatic compile-time optimizations that significantly improve the performance of MPC applications and incorporates the syntax of Python programming language to facilitate rapid application development. We demonstrate its usability and performance on various bioinformatics tasks showing up to 3–4 times increased speed over the existing pipelines with 7-fold reductions in codebase sizes.

Genomic data leaks are irreversible. Leaked DNA cannot be changed, stays disclosed indefinitely, and affects the owner's family members as well. The recent large-scale genomic data collections [1], [2] render the traditional privacy protection mechanisms, like the Health Insurance Portability and Accountability Act (HIPAA), inadequate for protection against the novel security attacks [3]. On the other hand, data access restrictions hinder important clinical research that requires large datasets to operate [4]. These concerns can be naturally addressed by the employment of privacy-enhancing technologies, such as a secure multiparty computation (MPC) [5]–[10]. Secure MPC enables computation on data without disclosing the data itself by dividing the data and computation between multiple computing parties in a distributed manner to prevent individual computing parties from accessing raw data. MPC systems are being increasingly adopted in fields that operate on sensitive datasets [11]–[13], such as computational genomics and biomedical research [14]–[22].