Evidence that a RecQ helicase slows senescence by resolving recombining telomeres.

RecQ helicases, including Saccharomyces cerevisiae Sgs1p and the human Werner syndrome protein, are important for telomere maintenance in cells lacking telomerase activity. How maintenance is accomplished is only partly understood, although there is evidence that RecQ helicases function in telomere replication and recombination. Here we use two-dimensional gel electrophoresis (2DGE) and telomere sequence analysis to explore why cells lacking telomerase and Sgs1p (tlc1 sgs1 mutants) senesce more rapidly than tlc1 mutants with functional Sgs1p. We find that apparent X-shaped structures accumulate at telomeres in senescing tlc1 sgs1 mutants in a RAD52- and RAD53-dependent fashion. The X-structures are neither Holliday junctions nor convergent replication forks, but instead may be recombination intermediates related to hemicatenanes. Direct sequencing of examples of telomere I-L in senescing cells reveals a reduced recombination frequency in tlc1 sgs1 compared with tlc1 mutants, indicating that Sgs1p is needed for tlc1 mutants to complete telomere recombination. The reduction in recombinants is most prominent at longer telomeres, consistent with a requirement for Sgs1p to generate viable progeny following telomere recombination. We therefore suggest that Sgs1p may be required for efficient resolution of telomere recombination intermediates, and that resolution failure contributes to the premature senescence of tlc1 sgs1 mutants.

Identifying wrist fracture patients with high accuracy by automatic categorization of X-ray reports.

The authors performed this study to determine the accuracy of several text classification methods to categorize wrist x-ray reports. We randomly sampled 751 textual wrist x-ray reports. Two expert reviewers rated the presence (n = 301) or absence (n = 450) of an acute fracture of wrist. We developed two information retrieval (IR) text classification methods and a machine learning method using a support vector machine (TC-1). In cross-validation on the derivation set (n = 493), TC-1 outperformed the two IR based methods and six benchmark classifiers, including Naive Bayes and a Neural Network. In the validation set (n = 258), TC-1 demonstrated consistent performance with 93.8% accuracy; 95.5% sensitivity; 92.9% specificity; and 87.5% positive predictive value. TC-1 was easy to implement and superior in performance to the other classification methods.