LAB NEWS
Feb 9, 2020

Alireza Nasiri, Jing Jin and Steph-Yves Louis Win Big Data Award

We congratulate our CSE Ph.D. students: Alireza Nasiri, Jing Jin and Steph-Yves Louis from the Machine Learning and Evolution Lab directed by Dr. Hu for winning the second place and $2,000 prize in the Big Data Health Science Case Competition organized by the UofSC National Big Data Health Science Conference, Columbia, SC February 7th-9th 2020. The objective of the case competition is to develop a machine learning or data analytic tool that will accurately identify the culprit chemical for a given patient. The provided dataset has a list of 311 chemicals and 79 corresponding signs and sympotoms and a list of ~1,000,000 patient records. There are total 19 teams from around the nation. The Big Data Case Competition is intended to provide enthusiastic teams of graduate and senior undergraduate students with the opportunity to apply their knowledge to the analysis of big datasets in healthcare. Participating teams will be given the same business problem, datasets and access to software to solve a challenging problem using a data analytics approach. Each participating team will analyze the case and data sets provided by the sponsoring company. During this period, the teams must identify the most prominent issues facing the company and develop a strategy to address these key issues in order to help companies achieve their goals. Team members will work together to present their methods and analysis results at the Big Data Health Science Conference. A panel of industry and academic experts judge the presentations based on teams’ use of the full analytics process, from framing the problem to methodology selection, data use, model building and innovation.




Feb 2, 2020

Dr. Jianjun Hu won the Research Progress Award

Dr. Jianjun Hu won the College of Engineering and Computig's 2019 Research Progress Award. This award recognizes particularly distinguished performance and accomplishments in the area of research by a tenured or tenure-track faculty member at the rank of Associate Professor in the College of Engineering and Computing.




December 17, 2019

Magellan Scholarship News

We congratulate the following Computer Science undergraduate students for receiving a Magellan Scholar Award for Spring 2020. Shayon Ghoshroy: Deep Residual Learning for Computational Identification of Amino Acids Christian Loftis: Genetic programming based Symbolic Regression for Material Thermal Conductivity Prediction


October 15, 2013, the USC Material Doping Databank is online. STEMIDB:STEM Imaging Database. It is a repository for depositing and retrieval of nanoscale images obtained by using scanning transmission electron microscopy (STEM). This is a collaboration prject with Professor Thomas Vogt of the Department of Chemistry at USC May 11, 2012. Logistic Regression based Ensemble server for protein localization prediction (yeast proteins) is online. visit LR ensemble Server


December 12, 2011. Ananda Mondal received his Ph.D. His dissertation is: NETWORK BASED PREDICTION OF PROTEIN LOCALIZATION USING DIFFUSION KERNEL.


Jan 8, 2010, Lewis Cawthorne won the University magellan Scholarship on developing text mining system for protein sorting motif analysis.


Stephanie won the University magellan Scholarship on developing webservers for protein sorting databases.


Aug 1, 2009. Dr. Hu was awarded NSF Career Award for project: Computational Analysis and Prediction of Genome-Wide Protein Targeting Signals and Localization.
This is a CAREER award to support the research of Dr. Jianjun Hu, in the Department of Computer Science and Engineering at University of South Carolina. Dr. Hu is a second-year, tenure-track Assistant Professor. A typical cell has a size of only 10 microns while it contains about a billion proteins. How these proteins are transported from their synthesis sites to their target locations within or outside of the cell is still not well understood. Experiments showed that translocation of nascent proteins are usually guided by postal code-like targeting signals encoded within the amino acid sequences of proteins. Genome-wide identification and decoding of these so-called molecular zip codes are fundamental to comprehensive understanding of the cell. Experimentally identifying protein targeting signals is labor-intensive. Computational prediction of targeting signals is still a big challenge due to their low conservation at the amino acid level. Currently, no de novo discovery algorithm is available for identifying new protein targeting signals. Also missing are appropriate models and algorithms for comparing these signals. This grant is 1) investigating novel computational algorithms for de novo discovery of new protein targeting signals; 2) developing models and algorithms for representing, detecting, and comparing targeting signals and 3) developing a protein functional network-based integrative algorithms for protein localization prediction. A transformative result of these studies will be a sequence encoding scheme based on amino acid indexes. This scheme will convert protein sequences into sequences of amino acid groups (AAGs) such that conserved patterns can be represented, modeled and discovered. Finally, protein function networks will be derived from models of protein localization prediction. With this research, computational identification and decoding of genome-wide protein targeting signals and precise protein localization predication will greatly enhance the understanding of how proteins are assembled in a cell. Tools developed during this project will be made available on the lab website: http://mleg.cse.sc.edu As a part of his CAREER grant, Dr. Hu will conduct short-term projects and student-run seminars to bring undergraduates into the bioinformatics research. A special effort will be made to change the perception that computer science is debugging code, as perceived by many high-school students. A novel computer game will be employed to show how bioinformatics addresses real-world problems. This will raise the public and especially the awareness and interest of K-12 students in bioinformatics. Students in the NSF STARTS Alliance program at the University of South Carolina will be targeted for students. Mini programming problems with a bioinformatics background will be developed for lower-level college students so that they will be exposed to bioinformatics early in their introductory programming courses. This project will also develop bioinformatics web services for de novo discovery, comparison, and retrieval of protein targeting signals and precise protein localization prediction.
For more detail see: Visit NSF Award page now.


July 28, 2009. AAEnrich: Identifying enriched amino acid compostions with special physichemical properties for protein sorting motifs
This server will allow one to identify whether there are over-representation of a special class of amino acids exisitng in a group of protein sequences. E.g. secretary sorting motifs will be shown to have over-represented hydrophobic and polar amino acids.!
Visit AAEnrich server now.


June 1, 2009. Stephanie Henrichs joins us for her 2-month summer REU intern.
She will be working on protein protein interaction for sorting motif analysis. She has programming experience in java, c/C++ and background knowledge in molecular biology. welcome!


March 2009. Annda Mondal joined our lab!
He will work on data mining and machine learning in virtual screening problem. He has been working on a pipeline for virtual screening 8 millions ligands for a therepeutic target of breast cancer.


January 2009. Dr. Hu's new book published: Genetic Programming and Creative Design of Mechatronic Systems, China Machine Press, Beijing China.
This book introduces the techniques of using evolutionary genetic programming for engineering design innovation.This book is published in Chinese.


2009.1 Joint paper Disease diagnosis using multi-microarray datasets published in BMC bioinformatics.



August, 2008: Fan and Jia join the MLEG lab, Welcome!
June 18, 2008. New software Online metaP: meta-server for protein localization prediction