A voluntary program to enable improved predictive modeling for faster drug discovery
SynSight is Syngene’s data consortium, comprising a group of consenting clients working towards enabling greater learning and creating superior mathematical models for more informed decisions in their projects. It is a not-for-profit scientific program that provides a common platform for its members to share their data anonymously and safely to enable faster drug discovery.
Through SynSight, we envision democratizing data to develop better analytics and domain applications for drug discovery research and development. SynSight allows its members to pool their resources to leverage data that aids design and improves decision-making, thereby shortening the design-make-test-analyze (DMTA) cycle.
Participation in the SynSight is voluntary. Being a part of the consortium allows access to larger datasets, consequently improving the quality of the built models.
Ensuring data confidentiality and leak prevention
SynSight ensures confidentiality by applying solutions such as anonymized and federated learning to grant anonymity to the data in the consortium. The anonymized learning solution embeds data into ‘fingerprints’ or ‘features,’ which are then used to build the models. The solution makes reconstruction of the parent data impossible while maintaining statistical accuracy at the same time.
Where possible, we use federated learning to avoid aggregating data and to preserve data confidentiality. No data (raw or anonymized) will be available in the consortium. At any time in model-building, SynSight will not use or mix the parent data provided by the contributing members. Rather, it will use only cleaned, anonymized data or models.
To know more about our Data Consortium, Download the Brochure: