About
Protein crystallization has long been a labor-intensive bottleneck in protein structure determination. Over the last few decades, the adoption of innovative lab automation, including liquid handlers and imagers, has enabled scientists to focus more on experimental design, greatly increasing their throughput and ultimately resulting in significantly more data.
As a result, scientists found themselves spending more and more time evaluating images from their crystallization experiments. To alleviate this burden, Formulatrix then embarked on developing AI-based methods to fully automate drop scoring and crystal identification. In 2020, we integrated MARCO, a scientist-led and Google-developed auto-scoring model, into the Rock Maker.
Although MARCO was reasonably effective, it struggled to identify crystals mixed with other classes in a drop. To improve this, Formulatrix developed Sherlock, an advanced AI auto-scoring model with the ability to identify crystals that are present with some other classes in a drop.
Sherlock, trained on larger and more diverse datasets and equipped with new features, also outperforms MARCO in its agreement with scientist-scored images. Rock Maker v4 now includes both Sherlock and MARCO, offering users the choice of auto-scoring models in scoring their crystallization experiments.
Join us in this webinar to learn more about how Sherlock can enhance the scoring of your crystallization experiments, saving you time and providing you confidence in your results.
About the Speaker:
Hafsa Iftikhar is currently working at Formulatrix as an application scientist for protein crystallization. She received her Ph.D. in Biochemistry and Structural Biology with research focused on the Identification of new potential inhibitors targeting HCV and SARS-CoV-2. Before joining Formulatrix, she worked as a research fellow at Helmholtz Centre for Infection Research (HZI), Germany, and associate researcher at Lahore University of Management Sciences (LUMS), Pakistan.