A K Mitra, U K Mukherjee, T Harding, J S Jang, H Stessman, Y Li, A Abyzov, J Jen, S Kumar, V Rajkumar, B Van Ness

Multiple myeloma (MM) is characterized by significant genetic diversity at subclonal levels that play a defining role in the heterogeneity of tumor progression, clinical aggressiveness and drug sensitivity. Although genome profiling studies have demonstrated heterogeneity in subclonal architecture that may ultimately lead to relapse, a gene-expression based prediction program that can identify, distinguish and quantify drug response in subpopulations within a bulk population of myeloma cells is lacking. In this study, we performed targeted transcriptome analysis on 528 pre-treatment single-cells from 11 myeloma cell lines and 418 single-cells from 8 drug-naïve MM patients followed by intensive bioinformatics and statistical analysis for prediction of proteasome inhibitor (PI)-sensitivity in individual cells. Using our previously reported drug response GEP signature at the single-cell level, we developed an R Statistical analysis package available at https://github.com/bvnlab/SCATTome, SCATTome (Single Cell Analysis of Targeted Transcriptome), that restructures the data obtained from Fluidigm single-cell qRT-PCR analysis run, filters missing data, performs scaling of filtered data, builds classification models, and predicts drug response of individual cells based on targeted transcriptome using an assortment of machine learning methods. Application of SCATT should contribute to clinically relevant analysis of intra-tumor heterogeneity, and better inform drug choices based on sub-clonal cellular responses.Leukemia accepted article preview online, 29 December 2015. doi:10.1038/leu.2015.361.