MAR 06, 2018 08:00 AM PST

Performance assessment of RNA sequencing and expression arrays for transcriptome analysis in cancer research

  • Research Scientist, Bioinformatics and Modelling Group, Proteome and Genome Research Unit, Department of Oncology, Luxembourg Institute of Health
      Dr. Petr Nazarov is a Research Scientist in the Proteome and Genome Research Unit of the Department of Oncology at Luxembourg Institute of Health and an invited Assistant Professor at Luxembourg University. Dr. Nazarov produced over 30 publications devoted to data analysis and oncology. His primary expertise is biostatistics, machine learning and genomics. His recent work focuses on characterizing transcriptional variations in peritumoral tissue environment of pancreatic cancer, as well as tumors originated from lung, colon, skin and brain. His current research leverages deconvolution of transcriptomic signals from heterogeneous tumors to improve cancer classification.


    DATE: March 6, 2018
    TIME: 08:00am PST, 11:00am EST, 5:00pm CET

    RNA sequencing and expression arrays are transcriptomics techniques used to quantify transcribed genes and their isoforms. The first part of the talk will compare RNA-seq and array performance for the analysis of differences between normal lung epithelium tissue and squamous cell carcinoma lung tumors. High concordance for protein-coding gene expression was observed between the platforms. However, much more discrepant results were obtained for the detection of less abundant transcripts, including long non-coding RNAs and alternatively spliced variants. The second part of the talk will describe the method that can help combine results from different platforms and analyze data generated from bulk heterogeneous samples, called independent component analysis.

    In this webinar, the speaker will:

    • Describe the results of platform comparison for protein-coding and long non-coding genes
    • Explain the biases observed in RNA-seq and expression array platforms
    • Share insights on how deconvolution of transcriptomics data followed by machine learning can improve tumor classification


    For Research Use Only. Not for use in diagnostic procedures.

    Show Resources
    Loading Comments...