Identification of a 3-aminoimidazo[1,2-a]pyridine inhibitor of HIV-1 reverse transcriptase.

BACKGROUND: Despite the effectiveness of highly active antiretroviral therapy (HAART), there remains an urgent need to develop new human immunodeficiency virus type 1 (HIV-1) inhibitors with better pharmacokinetic properties that are well tolerated, and that block common drug resistant virus strains. METHODS: Here we screened an in-house small molecule library for novel inhibitors of HIV-1 replication. RESULTS: An active compound containing a 3-aminoimidazo[1,2-a]pyridine scaffold was identified and quantitatively characterized as a non-nucleoside reverse transcriptase inhibitor (NNRTI). CONCLUSIONS: The potency of this compound coupled with its inexpensive chemical synthesis and tractability for downstream SAR analysis make this inhibitor a suitable lead candidate for further development as an antiviral drug.

Multivariate Uncertainty in Deep Learning.

Deep learning has the potential to dramatically impact navigation and tracking state estimation problems critical to autonomous vehicles and robotics. Measurement uncertainties in state estimation systems based on Kalman and other Bayes filters are typically assumed to be a fixed covariance matrix. This assumption is risky, particularly for "black box" deep learning models, in which uncertainty can vary dramatically and unexpectedly. Accurate quantification of multivariate uncertainty will allow for the full potential of deep learning to be used more safely and reliably in these applications. We show how to model multivariate uncertainty for regression problems with neural networks, incorporating both aleatoric and epistemic sources of heteroscedastic uncertainty. We train a deep uncertainty covariance matrix model in two ways: directly using a multivariate Gaussian density loss function and indirectly using end-to-end training through a Kalman filter. We experimentally show in a visual tracking problem the large impact that accurate multivariate uncertainty quantification can have on the Kalman filter performance for both in-domain and out-of-domain evaluation data. We additionally show, in a challenging visual odometry problem, how end-to-end filter training can allow uncertainty predictions to compensate for filter weaknesses.

A 3D Probabilistic Deep Learning System for Detection and Diagnosis of Lung Cancer Using Low-Dose CT Scans.

We introduce a new computer aided detection and diagnosis system for lung cancer screening with low-dose CT scans that produces meaningful probability assessments. Our system is based entirely on 3D convolutional neural networks and achieves state-of-the-art performance for both lung nodule detection and malignancy classification tasks on the publicly available LUNA16 and Kaggle Data Science Bowl challenges. While nodule detection systems are typically designed and optimized on their own, we find that it is important to consider the coupling between detection and diagnosis components. Exploiting this coupling allows us to develop an end-to-end system that has higher and more robust performance and eliminates the need for a nodule detection false positive reduction stage. Furthermore, we characterize model uncertainty in our deep learning systems, a first for lung CT analysis, and show that we can use this to provide well-calibrated classification probabilities for both nodule detection and patient malignancy diagnosis. These calibrated probabilities informed by model uncertainty can be used for subsequent risk-based decision making towards diagnostic interventions or disease treatments, as we demonstrate using a probability-based patient referral strategy to further improve our results.

Functional analysis of a conserved region of the baculovirus envelope fusion protein, LD130.

The envelope fusion protein from a baculovirus pathogenic for Lymantria dispar was characterized. N-terminal sequence analysis determined that it was cleaved downstream of predicted signal peptide and furin cleavage motifs. Mutation of the furin motif resulted in a protein that was not cleaved and did not mediate fusion. Mutagenesis of three charged amino acids in a conserved sequence with the features of a fusion peptide resulted in significant reduction of the ability of the constructs to mediate fusion. None of the mutations inhibited transport of the proteins to the cell surface. In addition, the mutations of the predicted fusion peptide region yielded no inhibition of cleavage. No difference in cleavage was detected between constructs expressed in Spodoptera frugiperda or L. dispar cells.

Transcriptional mapping of two genes encoding baculovirus envelope-associated proteins.

Genes encoding two representatives of the LD130 family of baculovirus envelope-associated proteins were transcriptionally mapped. These included ld130, which encodes a low pH-induced envelope fusion protein of the Lymantria dispar multinucleocapsid nucleopolyhedrovirus, and op21, which is related to ld130 but is encoded by Orgyia pseudotsugata MNPV and appears to lack an envelope fusion activity. The size and temporal expression of mRNA of both genes were examined by Northern blot analysis of RNA extracted from infected cells at selected timepoints. In addition, 5' rapid amplification of cDNA ends (RACE) in combination with DNA sequence analysis was used to map the start sites of mRNA. Ld130 predominantly utilized its early promoter at 24 h post-infection but by 72 h post-infection ld130 expression was almost exclusively from its late promoter. In contrast, op21 was expressed predominantly from its early promoter throughout the timecourse, even though a consensus late promoter sequence was present within 100 bp of the translation start codon. A significant fraction of late transcripts that mapped to op21 were spliced transcripts originating in the op18 gene region. The 3' termini of the transcripts were also mapped using 3' RACE.