Genomic Bioinformatics Laboratory

Genomic Bioinformatics laboratory

Mali Salmon-Divon’s lab

General Research Interests

Many of the challenges in genomics derive from the informatics needed to analyze and store the huge amount of data become available from microarray and sequencing technologies

Modern genomic analysis involves the processing of high-throughput experimental data, and integration of data coming from different sources so as to understand genetic regulation and function. Utilization of these techniques requires the development of new algorithms and tools for effective storage, manipulation, and visualization of genomic data.

My group uses and develop advanced computational and bioinformatic tools so as to bring high-end genomics technologies from the bench to the bedside. By employing genome analysis methods to realize the potential of personalized medicine and by collaborating closely with experimental scientists and clinicians, my laboratory is involved in the interpretation of the functional effects of genomic, transcriptomic and epigenetics variations, and the translation of these discoveries into medical practice.

Examples of research topics include:

Overall characterization of the miRNA landscape in health and disease

MicroRNAs are small non-coding RNA molecules that regulate gene expression post-transcriptionally and either repress protein translation or induce mRNA cleavage and degradation. The interplay between abnormalities in protein-coding genes and miRNAs has been among the most exciting discoveries in oncology over the past decade. miRNAs have been shown to be crucial for cancer initiation, progression, and treatment response or resistance. In my laboratory, we study the role that miRNAs play in normal biological processes and in the pathogenesis of several diseases and malignancies (including AL amyloidosis, breast cancer, and medulloblastoma).

Genetic characterization of brain tumors in children for improving disease classification and detecting a minimal-residual disease

Medulloblastoma (MB) is the most common malignant brain tumor in children, and one of the deadliest. Patients with tumor dissemination have a low recovery rate, but imaging and cytology are not always sufficiently sensitive to detect early metastatic spread. Due to the importance of an early detection of metastatic disease and the lack of efficient clinical solutions, new diagnostic tools are needed for MB. Importantly, MB is divided into four tumor subgroups—each with a distinct clinical, biological, and genetic profile—which represent distinct molecular entities. Thus, targeted treatment should be designed according to MB class and the ability to easily differentiate between MB classes is, therefore, important for clinical applications. To meet these needs, we combine bioinformatics and machine learning techniques with experimental approaches to improve the characterization and classification of MB subgroups and to improve the detection and monitoring of minimal residual disease in a personalized manner

Application of the Benford Law in biology

The reduction in sequencing costs has led to an unprecedented trove of gene expression data from diverse biological systems. Subsequently, principles from other disciplines can now be examined using this high-throughput transcriptomic information. One such principle is the Benford Law, which states that in numerical data, the proportion of numbers beginning with any given digit is not uniform but, rather, skewed, with 1 being the most common digit and 9 being the rarest. Whereas the Benford Law could previously be evaluated only by using data-rich systems, we are today able to test whether gene expression data has a Benford distribution, and we can determine how deviation from this distribution can be utilized to detect the tendency of a gene to have specific roles in the examined tissue. In addition, we study whether the Benford test can detect cancer driver genes in a personalized manner and identify cell types and cell origin. This ability could be used in various medical applications, including to identify metastasis origin, or to identify rare cells within a given population. These could contribute to cancer diagnosis, treatment, and regenerative medicine.