Back

Case Study of Single-cell Protein Activity Based Drug Prediction for Precision Treatment of Cholangiocarcinoma

Obradovic, A.; Tomassoni, L.; Yu, D.; Guillan, K.; Souto, K.; Fraser, E.; Bates, S.; Drake, C. G.; Saenger, Y.; Cruz, F. D.; Kung, A.; Califano, A.

2022-03-02 cancer biology
10.1101/2022.02.28.482410 bioRxiv
Show abstract

Cholangiocarcinoma is a rare, aggressive malignancy with limited treatment options, due to a paucity of actionable mutations and low response to immune checkpoint inhibitors. Furthermore, its extreme heterogeneity prevents identification of actionable dependencies from bulk-tissue profiles. To address these challenges, we introduce a highly generalizable, single-cell framework for the mechanism-based prioritization of drugs to treat rare, highly heterogeneous tumors. Analysis of transformed cells, accounting for only 10% of a cholangiocarcinoma patient biopsy revealed three molecularly-distinct subpopulations, predicted to be sensitive to four drugs by regulatory network-based analysis. Validation in a low-passage, patient-derived xenograft (PDX) from the same patient confirmed tumor growth rate control by two of these drugs (plicamycin and dacinostat) and further validated predicted subpopulation-specific effects, suggesting they may represent promising candidates for follow-up clinical trials, either alone or in combination with current standard-of-care chemotherapies. The proposed approach can be generalized to elucidate complementary dependencies of rare, heterogeneous tumors, at the single cell level.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.