Mapping scientists' career trajectories in the survey of doctorate recipients using three statistical methods.

This paper investigates to what extent there is a 'traditional' career among individuals with a Ph.D. in a science, technology, engineering, or math (STEM) discipline. We use longitudinal data that follows the first 7-9 years of post-conferral employment among scientists who attained their degree in the U.S. between 2000 and 2008. We use three methods to identify a traditional career. The first two emphasize those most commonly observed, with two notions of commonality; the third compares the observed careers with archetypes defined by the academic pipeline. Our analysis includes the use of machine-learning methods to find patterns in careers; this paper is the first to use such methods in this setting. We find that if there is a modal, or traditional, science career, it is in non-academic employment. However, given the diversity of pathways observed, we offer the observation that traditional is a poor descriptor of science careers.

Leak or link? the overrepresentation of women in non-tenure-track academic positions in STEM.

This paper examines gender variation in departures from the tenure-track science, technology, engineering, and math (STEM) academic career pathway to non-tenure-track academic careers. We integrate multiple data sources including the Survey of Earned Doctorates and the Survey of Doctorate Recipients to examine longitudinal career outcomes of STEM doctorate women. We consider three types of careers after receipt of a PhD: academic, academic non-tenure-track, and non-academic positions. We find that STEM women are more likely to hold academic non-tenure-track positions, which are associated with lower job satisfaction and lower salaries among men and women. Explanations including differences in field of study, preparation in graduate school, and family structure only explain 35 percent of the gender gap in non-tenure-track academic positions.

Reconstruction of clonal trees and tumor composition from multi-sample sequencing data.

MOTIVATION: DNA sequencing of multiple samples from the same tumor provides data to analyze the process of clonal evolution in the population of cells that give rise to a tumor. RESULTS: We formalize the problem of reconstructing the clonal evolution of a tumor using single-nucleotide mutations as the variant allele frequency (VAF) factorization problem. We derive a combinatorial characterization of the solutions to this problem and show that the problem is NP-complete. We derive an integer linear programming solution to the VAF factorization problem in the case of error-free data and extend this solution to real data with a probabilistic model for errors. The resulting AncesTree algorithm is better able to identify ancestral relationships between individual mutations than existing approaches, particularly in ultra-deep sequencing data when high read counts for mutations yield high confidence VAFs. AVAILABILITY AND IMPLEMENTATION: An implementation of AncesTree is available at: http://compbio.cs.brown.edu/software.