A Bayesian network approach to feature selection in mass spectrometry data

Background  

Time-of-flight mass spectrometry (TOF-MS) has the potential to provide non-invasive, high-throughput screening for cancers and other serious diseases via detection of protein biomarkers in blood or other accessible biologic samples. Unfortunately, this potential has largely been unrealized to date due to the high variability of measurements, uncertainties in the distribution of proteins in a given population, and the difficulty of extracting repeatable diagnostic markers using current statistical tools. With studies consisting of perhaps only dozens of samples, and possibly hundreds of variables, overfitting is a serious complication. To overcome these difficulties, we have developed a Bayesian inductive method which uses model-independent methods of discovering relationships between spectral features. This method appears to efficiently discover network models which not only identify connections between the disease and key features, but also organizes relationships between features--and furthermore creates a stable classifier that categorizes new data at predicted error rates.     

Cell biology: Paper alert

A selection of interesting papers that were published in the two months before our press date in major journals most likely to report significant results in cell biology.     

Genetics and development: Paper alert

A selection of interesting papers that were published in the two months before our press date in major journals most likely to report significant results in genetics and development.     

The nucleotide addition cycle of RNA polymerase is controlled by two molecular hinges in the Bridge Helix domain

Background  

Cellular RNA polymerases (RNAPs) are complex molecular machines that combine catalysis with concerted conformational changes in the active center. Previous work showed that kinking of a hinge region near the C-terminus of the Bridge Helix (BH-H_C) plays a critical role in controlling the catalytic rate.     

Post-conception reproductive competition in cooperatively breeding common marmosets

Common marmosets are cooperatively breeding monkeys that exhibit high female reproductive skew. Subordinate females usually fail to breed as a consequence of ovulation suppression and inhibition of sexual behavior, and, even when they do breed, typically rear fewer infants than dominants. We evaluated possible mechanisms of post-conception reproductive competition by comparing hormonal profiles across pregnancy, pregnancy outcomes, infant survivorship, and behavior in laboratory-housed families containing one (N=9) or two (N=7) breeding females. Breeding females in plurally breeding groups did not exhibit well-defined dominance relationships and rarely engaged in escalated aggression with one another. No significant differences were found among singly breeding mothers, plurally breeding mothers, and plurally breeding daughters in urinary chorionic gonadotropin or estradiol sulfate concentrations during pregnancy, fetal biparietal diameter, frequency of spontaneous abortion, frequency of stillbirths, number of live-born infants per litter, or infant mortality rates. When females gave birth while another female in the family was pregnant, however, their infants were highly likely to be killed. The perpetrator was definitively identified in only one family, in which a pregnant female killed her daughter's infant. These results are consistent with observations of free-living common marmosets and suggest that breeding females do not regularly influence one another's pregnancy outcomes, but that they may commonly kill each other's infants, especially during their own pregnancy. Our findings further suggest that infanticide by breeding females may have selected for the evolution of reproductive restraint in subordinate female marmosets.