- Ubiquitin and proteasome pathways
- Cancer-related gene regulation
- Gene expression and cancer classification
- Protein Degradation and Inhibitors
- Protein Structure and Dynamics
- Error Correcting Code Techniques
- Algorithms and Data Compression
- Single-cell and spatial transcriptomics
- Wnt/β-catenin signaling in development and cancer
- RNA modifications and cancer
- Artificial Intelligence in Healthcare
- interferon and immune responses
- Immune Cell Function and Interaction
- Bioinformatics and Genomic Networks
- Lanthanide and Transition Metal Complexes
- Epigenetics and DNA Methylation
- Multiple Myeloma Research and Treatments
- Caching and Content Delivery
- Cell Image Analysis Techniques
- Neuropeptides and Animal Physiology
- Genomics and Chromatin Dynamics
- Receptor Mechanisms and Signaling
- Cancer-related Molecular Pathways
- Machine Learning in Healthcare
- Acute Myeloid Leukemia Research
Van Andel Institute
2017-2025
Grand Valley State University
2022-2024
Hope College
2015-2018
Wnt/β-catenin signaling is activated when extracellular Wnt ligands bind Frizzled (FZD) receptors at the cell membrane. Wnts FZD cysteine-rich domains (CRDs) with high affinity through a palmitoylated N-terminal “thumb” and disulfide-stabilized C-terminal “index finger,” yet how these binding events trigger receptor activation intracellular remains unclear. Here we report crystal structure of Frizzled-4 (FZD 4 ) CRD in complex palmitoleic acid, which reveals tetramer consisting two...
Studies in genetically 'identical' individuals indicate that as much 50% of complex trait variation cannot be traced to genetics or the environment. The mechanisms generate this 'unexplained' phenotypic (UPV) remain largely unknown. Here, we identify neuronatin (NNAT) a conserved factor buffers against UPV. We find Nnat deficiency isogenic mice triggers emergence bi-stable polyphenism, where littermates emerge into adulthood either 'normal' 'overgrown'. Mechanistically, is mediated by an...
Abstract Non-negative matrix factorization (NMF) is a popular method for analyzing strictly positive data due to its relatively straightforward interpretation. However, NMF has reputation as less efficient alternative the singular value decomposition (SVD), standard operation that highly optimized in most linear algebra libraries. Sparse single-cell sequencing assays, now feasible thousands of subjects and millions cells, generate matrices with tens non-negative transcript abundance entries....
Large, sparse matrices are common in a variety of applications scientific computing and machine learning, most commonly stored CSC, CSR, or COO format. Most extensions adaptations these formats focus on enabling faster computation through techniques such as blocking. While common, primarily take advantage sparsity nonzero patterns. In some cases, however, data is highly redundant, with relatively few unique values. We introduce two novel CSC that aim to exploit redundancy: 1)...
Angiogenesis is essential for cancer metastasis, thus the discovery and characterization of molecules that inhibit this process important. Thalidomide a teratogenic drug which known to angiogenesis effectively yet specific cellular targets its effect are not well known. We discovered CUL5 (previously identified as VACM-1), scaffold protein in E3 ligase complexes, involved thalidomide-dependent inhibition endothelial cell growth. Our results show human cells (HUVEC), decrease growth was...
Despite decades of research, acute myeloid leukemia (AML) remains a remarkably lethal malignancy. While pediatric AML (pAML) carries more favorable prognosis than adult AML, the past 25 years large clinical trials have produced few improvements in pAML survival. Nowhere is this evident patients carrying t(16;21)(p11;q22) translocation, which yields
Determining the appropriate rank in Non-negative Matrix Factorization (NMF) is a critical challenge that often requires extensive parameter tuning and domain-specific knowledge. Traditional methods for determination focus on identifying single optimal rank, which may not capture complex structure inherent real-world datasets. In this study, we introduce novel approach called Residual Sensitivity to Intial Conditions (RSIC) suggests potentially multiple ranks of interest by analyzing...
Expression of the VACM‐1/CUL5 gene, a subunit an E3 ubiquitin ligase complex, inhibits cellular proliferation in endothelial cell lines vitro through mechanism involving its post‐translational modification by ubiquitin‐like protein, NEDD8. Factors that regulate NEDD8ylation have not been investigated. We shown Resveratrol, naturally occurring phytoalexin, enhances antiproliferative effect VACM‐1/CUL5. In this study, we investigated Resveratrol on and NEDD8 expression localization Human...
Compressed Sparse Column (CSC) and Coordinate (COO) are popular compression formats for sparse matrices. However, both CSC COO general purpose cannot take advantage of any the properties data other than sparsity, such as redundancy. Highly redundant is common in many machine learning applications, genomics, often too large in-core computation using conventional storage formats. In this paper, we present two extensions to CSC: (1) Value-Compressed (VCSC) (2) Index- (IVCSC). VCSC takes high...