Predicting and Editing Genes
Details
We are excited to host our 3rd meetup of AI in Genomics!
In this event, Chanan (Evogene) and Shiran (TAU) will demonstrate how AI is used to predict plant's essential genes and how it is used in gene therapy (CRISPR).
Schedule:
18:00 - Gathering + Pizzas :-)
18:15 - Chanan Rubin (Core Technologies Director @Evogene)
"Discovery of novel essential plant genes for innovative herbicide development"
Weeds are a well-known problem in the agricultural world. They reduce crop yield, costing farmers around the world billions of dollars annually. Hand in hand with increased usage of herbicides, more and more weeds have developed resistance to currently available herbicides. The development of herbicides with novel modes of action is essential to overcome this rising resistance. One of the key elements in avoiding herbicide resistance is the design of small molecule inhibitors against novel targets.
In this talk, Chanan will present a machine learning approach to predict novel essential plant genes. Evogene has assembled over the years a massive amount of data on plants including gene expression patterns, gene-gene interactions, gene orthology groups etc. Using these resources, we calculated multiple gene features and used a training set of known essential genes to build a random forest classifier for predicting plant essential genes. This classifier is able to identify both known as well as novel essential genes. Notably, in-silico predictions were experimentally validated in the greenhouse, demonstrating the predictive capabilities of our system.
19:00 - Shiran Abadi (PhD Student @TAU)
"A machine learning approach for predicting CRISPR-Cas9 cleavage efficiencies"
The CRISPR system has generated much excitement in recent years as a genome editing technique. The modification of specific genes and genomic regions is enabled by designing CRISPR to target them according to sequence similarity to the DNA sequence. However, the efficacy of one design is not uniquely defined by exact identity to the target site, thus additional genes might be unintentionally modified.
In this talk, Shiran will present CRISTA (CRISPR Target Assessment), a novel algorithm within the machine learning framework that determines the probability of a genomic site to be cleaved by a given CRISPR design.