Clinical Persona released open-source package cross_svm on github. cross_svm is a faster, simpler version of the well-known libsvm Java software for Support Vector Machine learning.
Acute Myeloid Leukemia Challenge
Clinical Persona competed in DREAM Acute Myeloid Leukemia Outcome Prediction Challenge, a competition organized by MD Anderson Cancer Center, Rice University, SAGE Bionetworks and IBM Research. We finished second in predicting remission duration for patients undergoing chemotherapy. We finished third in predicting overall survival, and seventh in predicting remission status (yes/no) as a binary variable.
Highly-accessed manuscript on cross-validation
A peer-reviewed article "Cross-validation pitfalls when selecting and assessing regression and classification models", whose lead authors are Clinical Persona team members Damjan Krstajic and Ljubomir Buturovic, is the most accessed article ever published in Journal of Cheminformatics, as of December 29, 2016.
Bladder cancer poster presented at ASCO 2016
Clinical Persona team members co-authored an abstract entitled "Derivation of gene expression classifiers for the non-invasive detection of bladder cancer in the hematuria and recurrence surveillance populations". This work was discussed at the Poster Discussion Session at the American Society of Clinical Oncology annual meeting held in Chicago, June 3-7, 2016.
Clinical Persona machine learning algorithm powers new breast cancer recurrence assay
Clinical Persona developed a breast cancer recurrence algorithm used in CanAssist-Breast assay launched by OncoStem Diagnostics on November 30, 2016. The molecular test demonstrated clinically compelling performance in independent validation, and is an affordable alternative for many patients in the developing world.
Clinical Persona participated in the New England Journal of Medicine SPRINT Challenge
The SPRINT Data Analysis Challenge was about deriving novel clinically useful findings from SPRINT clinical trial which compared intensive and standard treatments for hypertension. We developed machine learning software SafeSPRINT which suggests whether a hypertension patient should receive the intensive treatment or standard.