Machine learning amplifies the effect of parental family history of Alzheimer's disease on list learning strategy.

Identification of preclinical Alzheimer's disease (AD) is an essential first step in developing interventions to prevent or delay disease onset. In this study, we examine the hypothesis that deeper analyses of traditional cognitive tests may be useful in identifying subtle but potentially important learning and memory differences in asymptomatic populations that differ in risk for developing Alzheimer's disease. Subjects included 879 asymptomatic higher-risk persons (middle-aged children of parents with AD) and 355 asymptotic lower-risk persons (middle-aged children of parents without AD). All were administered the Rey Auditory Verbal Learning Test at baseline. Using machine learning approaches, we constructed a new measure that exploited finer differences in memory strategy than previous work focused on serial position and subjective organization. The new measure, based on stochastic gradient descent, provides a greater degree of statistical separation (p = 1.44 × 10-5) than previously observed for asymptomatic family history and non-family history groups, while controlling for apolipoprotein epsilon 4, age, gender, and education level. The results of our machine learning approach support analyzing memory strategy in detail to probe potential disease onset. Such distinct differences may be exploited in asymptomatic middle-aged persons as a potential risk factor for AD.

Sleep-dependent improvement in visuomotor learning: a causal role for slow waves.

STUDY OBJECTIVES: Sleep after learning often benefits memory consolidation, but the underlying mechanisms remain unclear. In previous studies, we found that learning a visuomotor task is followed by an increase in sleep slow wave activity (SWA, the electroencephalographic [EEG] power density between 0.5 and 4.5 Hz during non-rapid eye movement sleep) over the right parietal cortex. The SWA increase correlates with the postsleep improvement in visuomotor performance, suggesting that SWA may be causally responsible for the consolidation of visuomotor learning. Here, we tested this hypothesis by studying the effects of slow wave deprivation (SWD). DESIGN: After learning the task, subjects went to sleep, and acoustic stimuli were timed either to suppress slow waves (SWD) or to interfere as little as possible with spontaneous slow waves (control acoustic stimulation, CAS). SETTING: Sound-attenuated research room. PARTICIPANTS: Healthy subjects (mean age 24.6 +/- 1.0 years; n = 9 for EEG analysis, n = 12 for behavior analysis; 3 women). MEASUREMENTS AND RESULTS: Sleep time and efficiency were not affected, whereas SWA and the number of slow waves decreased in SWD relative to CAS. Relative to the night before, visuomotor performance significantly improved in the CAS condition (+5.93% +/- 0.88%) but not in the SWD condition (-0.77% +/- 1.16%), and the direct CAS vs SWD comparison showed a significant difference (P = 0.0007, n = 12, paired t test). Changes in visuomotor performance after SWD were correlated with SWA changes over right parietal cortex but not with the number of arousals identified using clinically established criteria, nor with any sign of "EEG lightening" identified using a novel automatic method based on event-related spectral perturbation analysis. CONCLUSION: These results support a causal role for sleep slow waves in sleep-dependent improvement of visuomotor performance.

An ultraefficient affinity-based high-throughout screening process: application to bacterial cell wall biosynthesis enzyme MurF.

The authors describe the discovery of a new class of inhibitors to an essential Streptococcus pneumoniae cell wall biosyn-thesis enzyme, MurF, by a novel affinity screening method. The strategy involved screening very large mixtures of diverse small organic molecules against the protein target on the basis of equilibrium binding, followed by iterative ultrafiltration steps and ligand identification by mass spectrometry. Hits from any affinity-based screening method often can be relatively nonselective ligands, sometimes referred to as "nuisance" or "promiscuous" compounds. Ligands selective in their binding affinity for the MurF target were readily identified through electronic subtraction of an empirically determined subset of promiscuous compounds in the library without subsequent selectivity panels. The complete strategy for discovery and identification of novel specific ligands can be applied to all soluble protein targets and a wide variety of ligand libraries.

Kinase drug discovery by affinity selection/mass spectrometry (ASMS): application to DNA damage checkpoint kinase Chk1.

Kinase enzymes are involved in a vast array of biological processes associated with human disease; therefore, selective kinase inhibition by small molecules and therapeutic antibodies is an area of intense study. The authors show that drug candidates with immediate value for biological preclinical evaluation can be identified directly through ultra-efficient affinity screening of kinase enzymes and random compound mixtures. The screening process comprises sampling and trapping equilibrium binding between candidate ligands and protein in solution, followed by removal of unbound ligands via 3 rounds of ultrafiltration and direct identification of bound ligands by mass spectrometry. Evaluation of significant peaks is facilitated by automated integration and collation of the mass spectral data and import into custom software for analysis. One Chk1-selective ligand found by using this process is presented in detail. The compound is potent in both enzymatic and Chk1-dependent cellular assays, and specific contacts in the Chk1 active site are shown by X-ray crystallography.