Kernel testing compares cell states in single-cell data.
problem Comparing non-linear cell states in single-cell data.
method Kernel-based testing framework for non-linear distribution comparison.
result Identifies subtle population variations in cell states.
FL-Sailer enables federated learning for scATAC-seq data, reducing dimensionality and noise.
problem Privacy-preserving federated learning for ultra-high dimensional, sparse, and heterogeneous scATAC-seq data.
method FL-Sailer integrates adaptive leverage score sampling and an invariant VAE architecture.
result FL-Sailer converges to an approximate solution with bounded error, surpassing centralized methods.
Proposes network method to detect cancer biomarkers from DNA methylation patterns.
problem Detecting prognostic cancer biomarkers from epigenetic data.
method Network methodology to infer prognostic cancer biomarkers based on DNA methylation patterns.
result Identifies prognostic biomarkers termed 'network community oncomarkers'.
Efficiently learns HMMs across multiple cell types using spectral methods.
problem Learning parameters of large HMMs for comparative epigenomics is computationally challenging.
method Developed a latent variable model and an efficient spectral algorithm exploiting tree structure of hidden states.
result Provided sample complexity bounds and experimentally validated on nine human cell types.
Fast and cheaper next generation sequencing technologies will generate unprecedentedly massive and highly-dimensional genomic and epigenomic variation data. In the near future, a routine part of medical record will include the sequenced genomes. A fundamental question is how to efficiently extract genomic and epigenomi…
We present a nonparametric prior over reversible Markov chains. We use completely random measures, specifically gamma processes, to construct a countably infinite graph with weighted edges. By enforcing symmetry to make the edges undirected we define a prior over random walks on graphs that results in a reversible Mark…
Noise-filtering improves cancer drug sensitivity prediction.
problem Improving accuracy of cancer drug sensitivity predictions.
method Integrates numerical linear algebra and information retrieval techniques to filter out noisy cancer cell lines.
result Our approach yields the highest AUC on clinical trial data.
Machine learning integrates diverse biological data to understand complex phenomena.
problem Combining multiple data types to understand biological and medical phenomena.
method Developing effective models to integrate heterogeneous biological data.
result Machine learning can identify important features and predict outcomes from diverse biological data.
We present a nonparametric Bayesian method for disease subtype discovery in multi-dimensional cancer data. Our method can simultaneously analyse a wide range of data types, allowing for both agreement and disagreement between their underlying clustering structure. It includes feature selection and infers the most likel…
DeepDiff predicts differential gene expression from histone modifications using deep learning.
problem Predicting differential gene expression from histone modification signals, capturing combinatorial effects.
method Attention-based deep learning architecture with multiple LSTM modules and attention mechanisms.
result DeepDiff significantly outperforms state-of-the-art baselines for differential gene expression prediction.