Single Cell Analysis
The Frost lab has developed a number of bioinformatics methods for the analysis of single cell multi-omic data. These include techniques for gene set testing, cell typing, interaction analysis and receptor abundance estimation. Validation of these methods has included the translational application to tumor immunology datasets. Recent methods in this space include:
- Frost HR (2023). Reconstruction Set Test (RESET): a computationally efficient method for single sample gene set testing based on randomized reduced rank reconstruction error. bioRxiv
- Javaid A, Frost HR (2023). SPECK: An Unsupervised Learning Approach for Cell Surface Receptor Abundance Estimation for Single Cell RNA-Sequencing Data. Bioinformatics Advances (In Press)
- Schiebout CT, Frost HR (2022). CAMML with the Integration of Marker Proteins (ChIMP). Bioinformatics 38(23), 5206-5213
- Frost HR (2020). Variance-adjusted Mahalanobis (VAM): A fast and accurate method for cell-specific gene set scoring. Nucleic Acids Research 48(16), e94
Statistics/Applied Math
More abstract statistical and applied mathematics work provides an important foundation for our bioinformatics and biostatistics research. This work is generally motivated by the challenge of analyzing noisy and high-dimensional data. Recent techniques in this area include:
- Frost HR (2023). Iterative execution of discrete and inverse discrete Fourier transforms with applications for signal denoising via sparsification. arXiv
- Frost HR (2023). A quaternion model for single cell transcriptomics. bioRxiv
- Frost HR (2023). Eigenvector centrality for multilayer networks with dependent node importance. arXiv
- Frost HR (2022). Eigenvectors from Eigenvalues Sparse Principal Component Analysis (EESPCA). Journal of Computational and Graphical Statistics 31(2), 486-501
Cancer Genomics
We have created a range of techniques for the analysis of multi-omics cancer data at both gene and pathway levels. These methods have included pan cancer methods that model the association between somatic alterations and gene expression, methods that explore primary and metastatic lesions, methods that focus on tumor-immune interactions and methods that identify driver mutations. Recent work includes:
- Zheng X, Amos CI, Frost HR (2021). Pan-cancer evaluation of gene expression and somatic alteration data for cancer prognosis prediction. BMC Cancer. 21(1), 1-11
- Zheng X, Amos CI, Frost HR (2020). Cancer prognosis prediction using somatic point mutation and copy number variation data: a comparison of gene-level and pathway-based models. BMC Bioinformatics 21(1):1-19
- Kamal Y, Dwan D, Hoehn HJ., Sanz-Pamplona R, Alonso MH, Moreno V, Cheng C, Schell MJ, Kim Y, Felder SI, Rennert HS, Melas M, Lazaris C, Bonner JD, Rennert G, Gruber SB, Frost HR*, Amos CI*, Schmit SL* (2021). Tumor immune infiltration estimated from gene expression profiles predicts colorectal cancer progression. OncoImmunology 10(1), 1862529
- Kamal Y, Schmit SL, Hoehn HJ, Amos CI, Frost HR (2019). Transcriptomic differences between primary colorectal adenocarcinomas and distant metastases reveal metastatic colorectal cancer subtypes. Cancer Research 79(16):4227-4241
Tissue-Specific Gene Activity
Our group has developed computational approaches for quantifying the tissue-specific activity of protein-coding genes, as measured by gene expression or protein abundance, and used these tissue-specific scores to improve the statistical power and interpretation of genomic analyses. This has included work on tissue-specific gene set testing, use of normal tissue-specificity to improve analysis of cancer transcriptomics, and exploration of cell type-specificity.
- Frost HR (2023) Tissue-adjusted pathway analysis of cancer (TPAC). bioRxiv
- Frost HR (2021). Analyzing cancer gene expression data through the lens of normal tissue-specificity. PLOS Computational Biology 17(6), e1009085
- Frost HR (2018). Computation and application of tissue-specific gene set weights. Bioinformatics 34(17):2957-2964