Discovering genomic structure through learned neural architectures.
problem Decoding the complex, unknown structure of human genomes using deep learning.
method Developed a novel search algorithm to learn optimal architectures for genomic data.
result Architectures learned from RNA expression data predict gene regulatory structure and identify key sequence motifs.
Paper uses genome Markov structure for outlier detection and read classification.
problem Identifying outliers and classifying reads in genome databases.
method Applying second-order Markov models to triplet base distributions.
result Improved accuracy in outlier identification and read classification.
SVM and N-best algorithm classify microbial marker clades from genome sequences.
problem Classifying microbial clades from genome sequences, especially new species.
method Support vector machine (SVM) with N-best algorithm, time series feature extraction, random fragment generation, k-mer size selection.
result Recognition accuracy rates above 28% in top-1 candidate, above 91% in top-10 candidate.
Elastic co-clustering improves clustering of single-cell genomic data.
problem Improving clustering performance of single-cell genomic datasets.
method Elastic coupled co-clustering in an unsupervised transfer learning framework.
result Our algorithm significantly improves clustering performance over traditional methods.
Genomic models learn DNA sequences to predict functions.
problem Understanding complex genetic interactions.
method Training LLMs on DNA sequences to predict functions.
result gLMs can predict functions of DNA elements.
The paper predicts diseases using both clinical and genomics data.
problem Clinical predictions using genomics data are not common.
method Integrated clinical and genomics datasets, machine learning, Principal Component Analysis for feature selection.
result 73% accuracy in predicting 75 disease classes.
Prototype Matching Network (PMN) improves genomic TFBS prediction.
problem Predicting Transcription Factor Binding Sites (TFBSs) with hundreds of TFs as labels.
method Prototype Matching Network (PMN) that learns motif-like features and TF-TF interactions.
result PMN significantly outperforms baselines on a large TFBS dataset.
New method combines ensembling and regularization for genomic disease prediction.
problem Genomic diseases require accurate prediction and biomarker identification.
method Integrates regularization with ensembling techniques for high-dimensional binary classification.
result Identifies critical biomarkers overlooked by competing methods.
The increased affordability of whole genome sequencing has motivated its use for phenotypic studies. We address the problem of learning interpretable models for discrete phenotypes from whole genomes. We propose a general approach that relies on the Set Covering Machine and a k-mer representation of the genomes. We sho…
Dilated convolutions model long-distance genomic dependencies effectively.
problem Detecting regulatory elements from raw DNA with long-distance dependencies.
method Developed and used a novel dataset for dilated convolutional neural networks.
result Dilated convolutions are effective at modeling regulatory elements in the human genome.
Dr.S recommends cancer drugs based on genomic data.
problem Personalizing cancer treatments using genomic information.
method Machine learning to identify optimal drug-gene associations.
result Developed a Drug Recommendation System (Dr.S) for cancer cell lines.
With different genomes available, unsupervised learning algorithms are essential in learning genome-wide biological insights. Especially, the functional characterization of different genomes is essential for us to understand lives. In this book chapter, we review the state-of-the-art unsupervised learning algorithms fo…
PKB method uses pathway information for cancer sample classification.
problem Cancer genomic data's high dimensionality and limited sample sizes.
method Pathway-based Kernel Boosting (PKB) method integrating gene pathway information for sample classification.
result PKB method outperforms other methods and identifies relevant pathways.
Generates new human genomic sequences for LAI training.
problem Lack of accessible reference data sets for LAI.
method Class-conditional VAE-GAN to generate realistic sequences.
result Generated sequences improve LAI method performance.
Proposes FDR-corrected sparse CCA for neuroimaging and genomics.
problem High-dimensional datasets in neuroimaging and genomics make false discoveries a concern.
method FDR-corrected sparse canonical correlation analysis (CCA) for high-dimensional settings.
result The proposed method controls the FDR of canonical vectors in high-dimensional settings.
Develops a faster soybean genome clustering method combining spectral and vector quantization.
problem Clustering soybean whole genome sequences efficiently.
method Combines Spectral Clustering and Vector Quantization for computational efficiency.
result Significantly outperforms existing methods in cluster quality and time complexity.
Copula-based fusion improves breast cancer risk stratification.
problem Combining clinical and genomic risk scores using simple rules fails to capture their joint relationship.
method Used copulas to model the joint relationship between clinical and genomic risk scores.
result Copula-based fusion improves risk stratification, identifying subgroups with the worst prognosis.
Method extracts cancer signatures from genome data, reducing noise and variability.
problem Identifying stable cancer signatures from noisy genomic data.
method Applied statistical risk models from finance to cancer genome data, using NMF.
result Extracted signatures have lower variability and improved stability.
Private cancer prediction model trained on federated genomic data.
problem Train a private cancer prediction model on federated genomic data.
method Differentially private federated learning (FL) for genomic cancer prediction.
result Ranked 3rd in a competition for private cancer prediction.
TF-MoDISco finds transcription factor motifs from genomic data.
problem Identifying transcription factor motifs from genomic sequence data.
method Algorithm for motif discovery from basepair-level importance scores.
result Improved version v0.5.6.5 of TF-MoDISco.
SEISM tests neural network features for regulatory genomics.
problem Testing neural network features for regulatory genomics.
method Selective inference procedure for sequence motifs.
result Sampling under specific parameters characterizes composite null hypothesis.
Paper proposes scalable method for analyzing multi-omic data.
problem Integrating high-dimensional multi-omic data for cancer subtyping.
method Mixed graphical model approach using Birth-Death MCMC algorithm.
result Our method outperforms LASSO and standard BDMCMC in computational efficiency and model selection accuracy.
Neural network classifies liver cancer patients based on genomic data.
problem Classifying liver cancer patients into high-risk and low-risk groups.
method Data expansion using wavelet analysis, compression of wavelet coefficients, training a neural network model.
result The neural network model accurately classifies patients without survival time information.
In this paper we propose network methodology to infer prognostic cancer biomarkers based on the epigenetic pattern DNA methylation. Epigenetic processes such as DNA methylation reflect environmental risk factors, and are increasingly recognised for their fundamental role in diseases such as cancer. DNA methylation is a…
Measures DNA quality degradation effects.
problem Identifying degraded DNA sequence data.
method Novel quality quantification based on intentional degradation effects.
result Quantified measures of degradation can be used for multiple purposes.
iRF detects stable high-order interactions in genomics data.
problem Understanding high-order interactions in genomics data.
method Iterative Random Forest algorithm (iRF) for stable high-order interaction detection.
result iRF identifies stable high-order interactions with computational cost similar to Random Forest.
As the amount and complexity of genetic information increases it is necessary that we explore some efficient ways of handling these data. This study takes the "divide and conquer" approach for analyzing high dimensional genomic data. Our aims include reducing the dimensionality of the problem that has to be dealt one a…
Matrix completion has attracted significant recent attention in many fields including statistics, applied mathematics and electrical engineering. Current literature on matrix completion focuses primarily on independent sampling models under which the individual observed entries are sampled independently. Motivated by a…
Method computes embeddings for RNA-seq data without genome alignment.
problem No need for genome alignment for RNA-seq data analysis.
method RNN transforms kmers into 2D latent space for transcriptomic analysis.
result Captures DNA sequence similarity and abundance in latent space.
Understanding functional organization of genetic information is a major challenge in modern biology. Following the initial publication of the human genome sequence in 2001, advances in high-throughput measurement technologies and efficient sharing of research material through community databases have opened up new view…
Method corrects deep generative model likelihood scores for OOD detection.
problem Challenges in deploying neural networks on out-of-distribution data.
method Likelihood ratio method for deep generative models.
result Proposed method achieves state-of-the-art OOD detection performance.
Deep learning detects genetic interactions in type 2 diabetes.
problem Detecting genetic interactions in complex diseases like type 2 diabetes.
method Stacked Autoencoder for non-linear epistatic interactions.
result Deep learning can uncover missing heritability in complex diseases.
New algorithm detects changes in genomic data faster and more accurately.
problem Detecting changes in genomic data with constraints.
method Adapting a functional pruning technique to solve constrained changepoint detection problems.
result Log-linear time complexity algorithm achieves state-of-the-art accuracy.
fiBAG integrates multiplatform genomic data to identify disease markers.
problem Understanding complex mechanisms underlying human diseases from multiplatform genomic data.
method fiBAG uses Gaussian process models and Bayes factors to identify functional evidence and guide variable selection.
result fiBAG improves detection of disease-related markers compared to non-integrative methods.
Semi-supervised deep learning detects problematic reads for genome assembly.
problem De novo genome assembly is hindered by specific types of reads.
method Analysis of coverage graphs converted to 1D-signals using semi-supervised deep learning models.
result Semi-supervised deep learning models can detect problematic reads with minimal labeled data.
Optimization approach for efficient sampling in optical mapping for structural variant detection.
problem Efficient sampling strategy for structural variant detection using optical mapping.
method Developed an optimization approach using a hyper-geometric distribution and probabilistic concentration inequalities.
result Optimal sampling strategy requires sampling most chromosomal fragments to detect variants at high confidence with little biological material.
The study examines universal laws in finance, earthquakes, and genomes through inter-occurrence times.
problem Understanding the distribution of inter-occurrence times in various systems.
method Analyzes inter-occurrence times in finance, earthquakes, and genomes using q-statistics. result Identifies universal features and indices like q in these systems. PKB framework boosts genomic data analysis by integrating pathway knowledge.
problem Boosting discovery power and connecting new findings with biological mechanisms in genomic data.
method Pathway-based Kernel Boosting (PKB) framework integrating clinical and pathway information for prediction of various outcomes.
result PKB substantially outperforms other methods in predicting drug response and cancer survival.
Memory Matching Networks classify DNA sequences for protein binding sites.
problem Manual construction of DNA motifs is difficult due to their complexity.
method Memory Matching Networks (MMN) learn a dynamic memory bank of encoded motifs and match them to new sequences.
result MMN effectively classifies DNA sequences as protein binding or nonbinding sites.
Graphs represent gene segment organization, revealing complex interrelationships in a scrambled genome.
problem Understanding gene segment organization and interrelationships in a scrambled genome.
method Directed graphs representing gene segments and their relationships, with graph properties mapped to higher-dimensional space for analysis.
result Emerging star-like structures indicate complex interrelationships, including segments from multiple genes interleaving or overlapping.
An approach for learning ancestral causal relationships in high dimensions, validated on human genome-wide data.
problem Learning ancestral causal relationships in high-dimensional biological data.
method Supervised learning approach with discrete indicators treated as labels, scalable to large problems.
result The approach is highly effective and scalable to the human genome-wide setting, robust to perturbations of input information.
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…
Machine learning accurately diagnoses cancer from whole genome sequencing data.
problem Accurate cancer diagnosis at all stages.
method Novel MLAC (Machine Learning Against Cancer) method using next-gen RNA sequencing.
result Perfect precision, sensitivity, and specificity achieved for most tumor types.
Secure linear regression at speed of plaintext methods.
problem Secure multiparty linear regression and feature selection.
method Distributed algorithms combining geometric ideas.
result Efficient and secure genome-wide association studies.
A simple algorithm for GWAS estimating SNP effects.
problem Estimating main and epistatic effects of SNPs.
method Pairs of individuals considered instead of SNPs or pairs of SNPs, based on intuitive allele changes leading to phenotype differences.
result Weak dependence on the number of SNPs, strong on the number of individuals.
Proposes a copula-based model for multi-view clustering with directional dependency.
problem Challenges in integrating multi-source datasets with directional dependency.
method Copula-based multi-view clustering model accounting for directional dependence.
result Ignoring directional dependence negatively impacts clustering performance.
Bayesian analysis uncovers flux couplings in metabolic networks.
problem Uncertainty and unrealistic assumptions in traditional flux analysis methods.
method Introduces Bayesian metabolic flux analysis to model reactions probabilistically and infer flux distributions.
result Reveals informative flux couplings and more unobserved fluxes in metabolic networks.
Recent advances in high-throughput cDNA sequencing (RNA-Seq) technology have revolutionized transcriptome studies. A major motivation for RNA-Seq is to map the structure of expressed transcripts at nucleotide resolution. With accurate computational tools for transcript reconstruction, this technology may also become us…