Alva Rani James, PhD

Investigated and integrated multi-omics datasets via machine learning algorithms using a hodgepodge of tools such as Git/GitHub, Python (scikit-learn, pandas, jupyter notebook, et al), AWS, Docker, bioinformatics algorithms, R, R-Shiny, linear regression, NLP, BERTopic, workflow management languages (Snakemake), and project management systems (Jira) for the diverse research teams and collaborators.

• Responsible for co-leading cross-disciplinary initiatives to derive data insights from rare disease patients omics datasets by leveraging statistical, machine learning, and exploratory data analysis techniques
• Lead role in the data insights completeness analyses: Coordinated research projects to ensure timely reporting and interpretation of results to multidisciplinary research teams
• Applied generative AI to develop structured real-world data products for Pharmaceutical companies including Sanofi
• Developed reproducible pipeline for transcriptomics and metabolomics, and risk score for large patient datasets
• Build and deploy machine learning models for cross-disciplinary initiative, interpretation and explanation of predictions to various stakeholders

Identified the functions of non-coding RNAs in multiple cancer types by leveraging multi-omics data sets from the TCGA consortium and external evidence. Applied deep learning algorithms (Convolutional neural network, autoencoders) and bioinformatics tools for investigation and analysis of the data sets using libraries like TensorFlow, pandas, scikit-learn, NumPy, seaborn, and spark. Mentored master students for mini projects, reporting, and presentations.

• Developed reproducible pipeline using nextflow pipeline management system for variant effect prediction for UKbiobank exon seq data (github.com/HealthML/Nextflow-pipeline)
• Developed evidence-based functional prediction method for non-coding RNAs dysregulated in multiple cancer types

Analysed multiple omics cancer datasets arrived from hospitals across Switzerland.

• Developed multiple reproducible Bioinformatics pipelines using workflow management languages (RNA-seq pipeline) for NGS data sets (tetra byte) for three high-profile research projects.( The Tumor Profiler Study, Roche funded)
• Data integration and data structure management for data sets from multiple technologies including RNA-seq, Single-sequencing, proteomics, exome-seq, and whole-genome seq
• Versioning the codes and querying LabKey database for the metadata information of patients samples (Github, python APIs)
• Documented and developed pipelines for various audience
• Developed cancer-specific score for a given cancer type using Bayer score via R tool (hypoxia score)
• Sun grid engine and cloud computing - Clusters and Parallelization