Publications

Machine Learning Approach to Pattern Discovery in Oncogenomics

Machine Learning Approach to Pattern Discovery in Oncogenomics

In this paper, we present the development and validation of a novel algorithm for pattern discovery and its application in discovering biologically relevant genes and gene associations in unlabeled, and sparse genomic datasets. The strength of our method is not only in embracing the complexity of genetic architecture to identify distinct patterns embedded in a noisy background in a purely data-driven way but also, the ease with which it can be integrated into a comprehensive and flexible framework for data mining and actionable knowledge discovery that goes beyond human genetics and genomics.

Read more.

On the Road to Personalized Medicine: Discovery of Prognostic Combinatorial High-Order Interactions in Breast Cancer

On the Road to Personalized Medicine: Discovery of Prognostic Combinatorial High-Order Interactions in Breast Cancer

In this work, we developed and validated a systematic workflow that incorporates biomarker classifier and our Pattern Discovery EngineTM for accurate biomarker prediction and for the discovery of novel gene associations in search for novel, personalized strategies for combating breast cancer. Higher-order interactions were identified and validated based on published literature. Our methods provide novel insights into gene interaction patterns in breast cancer and deliver candidates for further study.

Read more.

Extracting Insights on the Dynamic Health-Disease Transitions in the Human Gut Microbiome

Extracting Insights on the Dynamic Health-Disease Transitions in the Human Gut Microbiome

Using the Pattern Discovery Engine™ coupled with a hypothesis-free approach, we analyzed a large dataset of 50 human microbiome samples, each with the relative abundance of the ~10,000 KEGG protein families. We identified 39 KEGGs that were significant in differentiating the disease states from each other and from healthy, with 9 of the KEGGs (out of 10K total) being most associated with a dynamic path from disease to health in the human-gut microbiome. Using our approach we were able to reduce the size of the dataset to be analyzed by three orders of magnitude.

Read more.

Development of Prognostic Gene Panels: Subtype Prediction & Risk Stratification in Breast Cancer

Development of Prognostic Gene Panels: Subtype Prediction & Risk Stratification in Breast Cancer

In this work, we have developed a robust and cost-efficient biomarker with a set of only six genes (Pattern BC06) that can predict both survival and subtypes with an accuracy of 86%. Our biomarker panel will allow an improved understanding of clinical implications of molecular subtypes and risk prediction that could help clinicians guide precision medicine, tailoring medical treatment to patients and their tumor characteristics.

Read more.

Detection, Diagnosis, and Localization of Pediatric Pneumonia Based on Pattern Detection in Chest Radiographic Images

Detection, Diagnosis, and Localization of Pediatric Pneumonia Based on Pattern Detection in Chest Radiographic Images

We have developed an ML-based decision support system to detect pneumonia in pediatric chest x-ray radiographs to expedite accurate diagnosis of the pathology. Our results suggest that a suite of targeted ML tools can be used to support multi-faceted diagnosis of childhood pneumonia in resource-constrained settings, compensating for the shortage of expensive equipment and specialists. Further research is necessary to determine the feasibility of applying this algorithm in a clinical setting and to test improved care and outcomes compared with current assessment of the pediatric community-acquired pneumonia.

Read more.