Summer REU Program

Follow

• 

  Applying a Named Entity Recognition Model to Chemical Patents

  WanXiang Chen, Zachery Nolan, Christina Tang, and Bridget T. McInnes

  Named Entity Recognition (NER) is an application of Natural Language Processing (NLP) that involves recording entities contained in a text excerpt and classifying them into predefined category types. Unstructured text, or text without a predefined format is converted into structured text that can serve as input to an NER model. Here, we use a pretrained NER model to extract and classify chemical entities from patents/ This streamlines the process of determining the number and types of chemical agents used in a specific patent. This information may be useful for researchers, pharmaceutical companies, and agencies such as the FDA.

• 

  Extracting PLGA synthesis parameters from scientific literature

  Bridget T. McInnes, Byron Hedden, and Nastassja Lewinksi

  Systematic Reviews are necessary to create an exhaustive summary of current evidence relevant to a specific research question. It is estimated that it takes six individuals over 1000 hours to complete a review. In this work, we evaluate automatically extracting synthesis parameters for Poly(lactic-co-glycolic acid) (PLGA) polymers

• 

  Decoding the Language of Chemistry: Exploring Reactions in Scientific Literature

  Grant Webb, August Moses, Christina Tang, and Bridget T. McInnes

  Natural Language Processing (NLP) offers promising potential for advancing chemistry research by exploring understudied topics, accelerating the research process, and reducing the costs of lab experimentation (1). For example, the U.S. military seeks to utilize NLP to replace toxic and limited resources, like Teflon, with more accessible alternatives. Our objective is to assess the accuracy of two models, Skip-Gram and BERT, after training them on chemistry data. Our work with Skip-Gram aims to replicate and enhance prior research results, while our work with BERT aims to highlight its Sub-Word method for analysis.

• 

  NLP@VCU: Crop Characteristic Extraction Framework

  Cora Lewis, Bridget McInnes, and Getiria Onsongo

  We developed a crop characteristic extraction framework. Starting from a custom SpaCy named entity recognition model, we added pre-trained word embeddings and a part-of-speech based entity expansion post-processing step. Then, we implemented an evaluation framework that functioned as a 5-fold cross validation wrapper for SpaCy custom training. Preliminary results showed improvement in the extraction framework after these additions.

• 

  Optimization of an Automated Algorithm for Analysis of Spontaneous Rhythmic Bladder Contractions During Urodynamics Testing

  Isabelle Pummill, Rui Li, Zachary Cullingsworth, Adam Klausner, and John Speich

• 

  Investigating Traction Forces of Breast Cancer K14+ Leader Cells in Tumor Organoids

  Ella Ramamurthy, Jessanne Lichtenberg, and Priscilla Hwang

• 

  Evaluation of Cell-Matrix Interactions in K14+ Leader Cells on CAF-Modulated Matrix

  Trey P. Redman, Jessanne Y. Lichtenberg, and Priscilla Y. Hwang

• 

  Assisting End-Users in Creating Chatbots by Improving Training Data

  Aparna Roy and Chris Egersdoerfer

• 

  Computer-Based Scaffolding In Computer Science Education

  Rebecca Trinh and Simone Levy

• 

  Exploiting the SARS-CoV-2 Spike Protein Components to Guide Molecular Level Entry of a BAG-1 Inhibitor in the Treatment of Breast and Lung Cancers

  Malak Yasin, Michael Peters, and Mo Jiang

  Chemoresistance of lung cancer cells is the primary reason as to why limitations occur with cancer treatments. A protein, known as BAG-1 is responsible for many cellular activities including cellular stress response, cell growth, and apoptosis (regulated cell death). When overexpressed, the protein has been linked to the anti-apoptotic behavior of cancer cells. BAG-1 can combine to heat shock proteins (HSPs), a family of helical molecular chaperones that are known to aid in the maturation of proteins, refolding, and degradation. This response plays a crucial role in the study of chemoresistance in cancer patients due to its detrimental nature. Prior, this combination was combated by using a synthesized poly-arginine linked peptide inhibitor alongside cell penetrating peptides (CPPs) through targeted binding domains. However, it has been found that the Spike protein of SARS-CoV-2 uses several small subdomains to efficiently bind to human epithelial cells at the nanomolar level. This study aims to focus on the binding complex of the BAG-1 and Spike proteins to form an anti-apoptotic inhibitor that can result in a potential specific binding mechanism for drug delivery of lung cancer treatments.