Professor, Department of Genetics and Biochemistry
Dr. F. Alex Feltus received a B.Sc. in Biochemistry from Auburn University in 1992, served two years in the Peace Corps in the Fiji Islands, and then completed advanced training in biomedical sciences at Vanderbilt and Emory. He has performed research in artificial intelligence, bioinformatics, cyberinfrastructure, high-performance computing, network biology, tumor biology, agrigenomics, genome assembly, systems genetics, paleogenomics, and bioenergy feedstock genetics. Currently, Feltus is Professor in Clemson University’s Dept. of Genetics & Biochemistry and CEO of Allele Systems LLC. Feltus has published numerous scientific articles in peer-reviewed journals, released open source software, and taught undergrad and PhD students in bioinformatics, biochemistry, and genetics. He is funded by multiple NSF grants and is engaged in tethering together extremely smart people from diverse technical backgrounds to propel genomics research from the Excel-scale into the Exascale.
Our group uses software engineering and computational biology techniques to make useful molecular discoveries in human and plant biological systems; We also engineer elastic advanced compute systems and technologies to run robust genomics workflows to enable small labs to perform innovative petascale computational biology. The lab also actively engaged in traditional PhD training and the development of a scalable asynchronous training platform for data-intensive computing including but not limited to computational biology. My lifetime research goal is to reveal the genomic mechanisms underlying phenotype expression. A core aspect of this approach to identify biomarkers that are able to group interesting biological states (e.g. normal kidney verses renal tumor somatic mutation and/or transcriptome profiles). Given that most traits are under control by complex cellular control systems, we always seek to identify sets of functionally interacting genes (biomarker systems) that discriminate between biological states. My group focuses on the transcriptome layer (RNA) of gene expression but we are always seeking methods to integrate data from other genome information orbitals.
A staple data construct of our lab is the gene co-expression network (GCN) where an edge represents a statistically significant RNA expression correlation between two gene products (network nodes). We are active developers of a GCN discovery software application called KINC that is able to identify condition-specific edges from mixed input gene expression matrices (GEMs) (Ficklin et al. ). KINC GCNs are made from GEMs in a bottom up approach where all gene pairs are tested for correlation. This approach is computationally intensive and is not be scalable to millions of samples. Further, traditional GCNs do not detect non-linear relationships missed by correlation tests and do not place genetic relationships in a gene expression intensity context. In response, we developed EdgeScaping (Husain and Feltus ), which constructs and analyzes the pairwise gene intensity network in a holistic, top down approach where no edges are filtered. EdgeScaping uses a novel technique to convert traditional pairwise gene expression data into an image based format and allows for exploring non-linear relationships between genes by leveraging deep learning image analysis algorithms. We have applied EdgeScaping to a human tumor expression profiles candidate biomarker systems that exhibit conventional and non-conventional interdependent non-linear behavior associated with brain specific tumor sub-types. Edgescaping is open source and available at [https://github.com/bhusain/EdgeScaping].
We have been mining RNA expression profiles for biomarker systems from many NIH projects including GTEx and TCGA. We are also leveraging open and protected PsychENCODE (Akbarian et al. ) and SPARK (Feliciano et al. ) datasets to better understand normal and aberrant brain expression patterns. Once we detect biomarker systems using the techniques described above, we try to understand the gene regulatory networks underlying those systems. We are focusing on detection and understanding biomarker systems for three specific biomedical phenotypes: intellectual disability (e.g. autism spectrum disorder — ASD), brain cancer, and renal cancer.
Genomics databases are swelling and larger compute systems are needed by my group and thousands of individual life science investigators. Soon, DNA sequencers will replace qPCR machines in research labs and everyone will need terascale/petascale compute systems. Towards this disruptive technological event on par with the roll out of molecular biology into labs in the 1980s, my group is actively engaged in several funded cyberinfrastructure projects: “CC*Data: National Cyberinfrastructure for Scientific Data Analysis at Scale (SciDAS).” NSF (Feltus PI); “RCN: Advancing Research and Education Through a National Network of Campus Research Computing Infrastructures – The CaRC Consortium” NSF (Feltus PI – Bottum Former PI); “Exposing the Potential of Information Centric Networks for the Life Sciences” Cisco Research (Feltus PI); “CC* NPEO: Toward the National Research Platform.” NSF(Smarr PI, Feltus End User); “DIBBs: EI: SLATE and the Mobility of Capability” NSF (Gardner PI, Feltus End User). Along with many others, I am linking these partnerships to help build larger democratized compute systems and scaling the training so people can actually use them. In addition to the workflow engineering outlined above, we are focusing efforts in these cutting edge three cyberinfrastructure areas: scaling out usage of Kubernetes based compute systems and moving genomics data from traditional data repositories into information centric network systems.
Cutcher-Gershenfeld J, Baker KS, Berente N, Berkman PA, Canavan P, Feltus FA, Garmulewicz A, Hutchins R, King JL, Kirkpatrick C, Lenhardt C, Lewis S, Maffe M, Mittleman B, Sampath R, Shin N, Stall S, Winter S, Veazey P. 2020. Negotiated sharing of pandemic data, models, and resources. Negotiation Journal. 36. doi: doi.org/10.1111/nejo.12340.
Hang Y, Aburidi M, Poehlman WL, Husain B, Hickman A, Feltus FA. 2020. Tissue specific brain gene co-expression networks reveal genes that sub-type human tumors. Sci Rep. 10. doi: 10.1038/s41598-020-73611-1.
Husain B, Hickman AR, Hang Y, Shealy BT, Sapra K and Feltus FA. 2020. NetExtractor: Extracting a cerebellar tissue gene regulatory network using differentially expressed high mutual information binary RNA profiles. G3 (Bethesda) 10: 2953-2963. doi: 10.1534/g3.120.401067.
Paterson AH, Kong W, Johnston RM, Nabukalu P, Wu G, Poehlman WL, Goff VH, Isaacs K, Lee TH, Guo H, Zhang D, Sezen UU, Kennedy M, Bauer D, Feltus FA, Weltzien E, Rattunde HF, Barney JN, Barry K, Cox TS and Scanlon MJ. 2020. The evolution of an invasive plant, Sorghum halepense L. (‘Johnsongrass’). Front Genet 11: 317. doi: 10.3389/fgene.2020.00317.
Pauly R, Ogle C, Mcknight C, Reddick D, Presley J, Shannigrahi S, Feltus FA. 2020. NDN-TR68: Utilizing NDN for domain science applications – a genomics example. NDN Technical Report.
Spoor S, Wytko C, Soto B, Chen M, Almsaeed A, Condon B, Herndon N, Hough H, Staton M, Wegrzyn J, Main D, Feltus FA, Ficklin S. 2020. Tripal and Galaxy: supporting reproducible scientific workflows for community biological databases. Database. doi: 10.1093/database/baaa032.
Targonski C, Bender MR, Shealy BT, Husain B, Smith MC, Paseman B, Feltus FA. 2020. Cellular state transformations using deep learning for precision medicine applications. Patterns. 1:6. doi: 10.1016/j.patter.2020.100087.
Casanova EL, Switala AE, Dandamudi S, Hickman AR, Vandenbrink J, Sharp JL, Feltus FA and Casanova MF. 2019. Autism risk genes are evolutionarily ancient and maintain a unique feature landscape that echoes their function. Autism Res 12: 860-869. doi: 10.1002/aur.2112.
Greene MA, Britt JL, Powell RR, Feltus FA, Bridges WC, Bruce T, Klotz JL, Miller MF and Duckett SK. 2019. Ergot alkaloid exposure during gestation alters: 3. Fetal growth, muscle fiber development and miRNA transcriptome. J Anim Sci 97: 3153-3168. doi:10.1093/jas/skz153.
Mills N, Bensman EM, Poehlman WL, Ligon WB 3rd and Feltus FA. 2019. Moving just enough deep sequencing data to get the job done. Bioinform Biol Insights 13: 1177932219856359. doi: 10.1177/1177932219856359.
Poehlman WL, Hsieh JJ and Feltus FA. 2019. Linking binary gene relationships to drivers of renal cell carcinoma reveals convergent function in alternate tumor progression paths. Sci Rep 9: 2899. doi: 10.1038/s41598-019-39875-y.
Spoor S, Cheng CH, Sanderson LA, Condon B, Almsaeed A, Chen M, Bretaudeau A, Rasche H, Jung S, Main D, Bett K, Staton M, Wegrzyn JL, Feltus FA and Ficklin SP. 2019. Tripal v3: an ontology-based toolkit for construction of FAIR biological community databases. Database (Oxford) doi: 10.1093/database/baz077.
Targonski C, Shearer CA, Shealy BT, Smith MC and Feltus FA. 2019. Uncovering biomarker genes with enriched classification potential from Hallmark gene sets. Sci Rep 9: 9747. doi: 10.1038/s41598-019-46059-1.
Casanova EL, Gerstner Z, Sharp JL, Casanova MF, Feltus FA. 2018. Widespread genotype-phenotype correlations in intellectual disability. Frontiers in Psychiatry 9: 535.
Falk T, Herndon N, Grau E, Buehler S, Richter P, Zaman S, Baker EM, Ramnath R, Ficklin S, Staton M, Feltus FA, Jung S, Main D, Wegrzyn JL. 2018. Growing and cultivating the forest genomics database, TreeGenes. Database pii: bay084
(Conference Proceedings) Shannigrahi S, Fan C, Papadopoulos C, Feltus FA. 2018. NDN-SCI for managing large scale genomics data. ICN 2018 Pages 204-205.
Roche KE, Weinstein M, Dunwoodie L, Poehlman WL, Feltus FA. 2018. Sorting five human tumor types reveals specific biomarkers and background classification genes. Scientific Reports 8: 8180.
Mills N, Feltus FA, Ligon III WB. 2018. Maximizing the performance of scientific data transfer by optimizing the interface between parallel file systems and advanced research networks. Future Generation Computer Systems 79: 190-198.
Dunwoodie LJ, Poehlman WL, Ficklin SP, Feltus FA. 2018. Discovery and validation of a glioblastoma co-expressed gene module.” Oncotarget 9: 10995-11008.
(Conference Proceedings) Russell T, Stealey M, Coposky J, Keller B, Castillo C, Idaszak R, Feltus FA. 2017. Distributing the iRODS catalog: A way forward. iRODS UGM 2017 Proceedings Page 35.
Livingstone III D, Stack C, Mustiga G, Rodezno D, Suarez C, Amores F, Feltus FA, Mockaitis K, Cornejo O, Motamayor JC. 2017. A larger chocolate chip – Development of a 15K Theobroma cacao L. SNP array to create high density linkage maps. Frontiers Plant Science 8: 2008.
(Conference Proceedings) Poehlman W, Rynge M, Desinghu B, Mills N, Feltus FA. 2017. OSG-KINC: High-throughput gene co-expression network construction using the open science grid. IEEE BIBM 2017 Proceedings Pages 1827-1831.
Ficklin SP, Dunwoodie LJ, Poehlman WL, Watson C, Roche K, Feltus FA. 2017. Discovering condition-specific gene co-expression patterns using Gaussian mixture models: A cancer case study. Scientific Reports 7: 8617.
Roche K, Feltus FA, Park JP, Coissieux MM, Chang C, Chan VBS, Bentires-Alj M, Booth BW. 2017. Cancer cell redirection biomarker discovery using a mutual information approach. PLoS One 12: e0179265.
Schmucker H, Park JP, Coissieux MM, Kwist K, Bentires-Alj M, Feltus FA, Booth BW. 2017. RNA expression profiling reveal differentially regulated growth factor and receptor expression in redirected cancer cells. Stem Cells and Development 26: 646-655.
Watts NA, Feltus FA. Big Data Smart Socket (BDSS): 2017. A tool that abstracts data transfer habits from end users. Bioinformatics 33: 627-628.
Poehlman WL, Rynge M, Branton C, Balamurugan D, Feltus FA. 2016. OSG-GEM: Gene expression matrix construction using the open science grid. Bioinformatics and Biology Insights 10: 133.
Wang Y, Ficklin SP, Wang X, Feltus FA, Paterson AH. 2016. Large-scale gene relocations following an ancient genome triplication associated with the diversification of core eudicots. PLoS One 11: e0155637.