In health care settings, patients who are physically proximate to other patients (co-presence) for a meaningful amount of time may have differential health outcomes depending on who they are in contact with. How to best measure this co-presence, however is an open question and previous approaches have limitations that may make them inappropriate for complex health care settings. Here, we introduce a novel method which we term “consistent co-presence”, that implicitly models the many complexities of patient scheduling and movement through a hospital by randomly perturbing the timing of patients’ entry time into the health care system. This algorithm generates networks that can be employed in models of patient outcomes, such as 1-year mortality, and are preferred over previously established alternative algorithms from a model comparison perspective. These results indicate that consistent co-presence retains meaningful information about patient-patient interaction, which may affect outcomes relevant to health care practice. Furthermore, the generalizabiity of this approach allows it to be applied to a wide variety of complex systems.

Entity resolution in settings with rich relational structure often introduces complex dependencies between co-references. Exploiting these dependencies is challenging - it requires seamlessly combining statistical, relational, and logical dependencies. One task of particular interest is entity resolution in familial networks. In this setting, multiple partial representations of a family tree are provided, from the perspective of different family members, and the challenge is to reconstruct a family tree from these multiple, noisy, partial views. This reconstruction is crucial for applications such as understanding genetic inheritance, tracking disease contagion, and performing census surveys. Here, we design a model that incorporates statistical signals, such as name similarity, relational information, such as sibling overlap, and logical constraints, such as transitivity and bijective matching, in a collective model. We show how to integrate these features using probabilistic soft logic, a scalable probabilistic programming framework. In experiments on real-world data, our model significantly outperforms state-of-the-art classifiers that use relational features but are incapable of collective reasoning.

A "bidirectional gene pair" comprises two adjacent genes whose transcription start sites are neighboring and directed away from each other. The intervening regulatory region is called a "bidirectional promoter." These promoters are often associated with genes that function in DNA repair, with the potential to participate in the development of cancer. No connection between these gene pairs and cancer has been previously investigated. Using the database of spliced-expressed sequence tags (ESTs), we identified the most complete collection of human transcripts under the control of bidirectional promoters. A rigorous screen of the spliced EST data identified new bidirectional promoters, many of which functioned as alternative promoters or regulated novel transcripts. Additionally, we show a highly significant enrichment of bidirectional promoters in genes implicated in somatic cancer, including a substantial number of genes implicated in breast and ovarian cancers. The repeated use of this promoter structure in the human genome suggests it could regulate co-expression patterns among groups of genes. Using microarray expression data from 79 human tissues, we verify regulatory networks among genes controlled by bidirectional promoters. Subsets of these promoters contain similar combinations of transcription factor binding sites, including evolutionarily conserved ETS factor binding sites in ERBB2, FANCD2, and BRCA2. Interpreting the regulation of genes involved in co-expression networks, especially those involved in cancer, will be an important step toward defining molecular events that may contribute to disease.