Max Ward

This paper introduces iterative sampling for RNA structure prediction, achieving faster and more comprehensive sampling than existing methods, and applies it to study cotranscriptional RNA folding.

mRNAfold improves mRNA design by co-optimizing stability, local structure, and codon choice, showing better in-solution stability and complex effects on protein production.

This paper evaluates deep learning methods for RNA secondary structure prediction, highlights generalization challenges, and discusses integration with traditional methods.

This paper introduces a training-free FFT-based method for efficient and interpretable epigenomic data compression, achieving high fidelity and speed compared to autoencoders.

The paper introduces a deep learning method using contrastive learning to predict gene expression from histone modifications without arbitrary thresholds, showing consistent performance across datasets.

This paper introduces a scalable differentiable folding algorithm for mRNA design, improving performance and enabling optimization via deep learning.

This study uses machine learning to identify novel shared targets in glioblastoma and osteosarcoma, revealing C2ORF72 as a potential therapeutic and diagnostic target.

This study uses classical statistical methods to identify survival-related genes in breast cancer, finding two novel targets and validating results through bootstrapping.

The paper introduces memerna, a fast RNA folding algorithm that incorporates coaxial stacking and sparsity, achieving the fastest exact RNA folding implementation among popular packages supporting this feature.

This paper introduces a differentiable version of McCaskill's RNA partition function algorithm, enabling gradient-based RNA sequence design for applications like mRNA vaccines.

The paper compares RNA design fitness functions and finds that maximizing probability outperforms minimizing ensemble defect.

The paper shows that deep learning models for RNA secondary structure prediction do not generalize across families, using rigorous cross-validation methods.