Scaling up molecular pattern recognition with DNA-based winner-take-all neural networks.

From bacteria following simple chemical gradients1 to the brain distinguishing complex odour information2, the ability to recognize molecular patterns is essential for biological organisms. This type of information-processing function has been implemented using DNA-based neural networks3, but has been limited to the recognition of a set of no more than four patterns, each composed of four distinct DNA molecules. Winner-take-all computation4 has been suggested5,6 as a potential strategy for enhancing the capability of DNA-based neural networks. Compared to the linear-threshold circuits7 and Hopfield networks8 used previously3, winner-take-all circuits are computationally more powerful4, allow simpler molecular implementation and are not constrained by the number of patterns and their complexity, so both a large number of simple patterns and a small number of complex patterns can be recognized. Here we report a systematic implementation of winner-take-all neural networks based on DNA-strand-displacement9,10 reactions. We use a previously developed seesaw DNA gate motif3,11,12, extended to include a simple and robust component that facilitates the cooperative hybridization13 that is involved in the process of selecting a 'winner'. We show that with this extended seesaw motif DNA-based neural networks can classify patterns into up to nine categories. Each of these patterns consists of 20 distinct DNA molecules chosen from the set of 100 that represents the 100 bits in 10 × 10 patterns, with the 20 DNA molecules selected tracing one of the handwritten digits '1' to '9'. The network successfully classified test patterns with up to 30 of the 100 bits flipped relative to the digit patterns 'remembered' during training, suggesting that molecular circuits can robustly accomplish the sophisticated task of classifying highly complex and noisy information on the basis of similarity to a memory.

Compiler-aided systematic construction of large-scale DNA strand displacement circuits using unpurified components.

Biochemical circuits made of rationally designed DNA molecules are proofs of concept for embedding control within complex molecular environments. They hold promise for transforming the current technologies in chemistry, biology, medicine and material science by introducing programmable and responsive behaviour to diverse molecular systems. As the transformative power of a technology depends on its accessibility, two main challenges are an automated design process and simple experimental procedures. Here we demonstrate the use of circuit design software, combined with the use of unpurified strands and simplified experimental procedures, for creating a complex DNA strand displacement circuit that consists of 78 distinct species. We develop a systematic procedure for overcoming the challenges involved in using unpurified DNA strands. We also develop a model that takes synthesis errors into consideration and semi-quantitatively reproduces the experimental data. Our methods now enable even novice researchers to successfully design and construct complex DNA strand displacement circuits.

Improving Computer-Aided Detection Using Convolutional Neural Networks and Random View Aggregation.

Automated computer-aided detection (CADe) has been an important tool in clinical practice and research. State-of-the-art methods often show high sensitivities at the cost of high false-positives (FP) per patient rates. We design a two-tiered coarse-to-fine cascade framework that first operates a candidate generation system at sensitivities  âˆ¼ 100% of but at high FP levels. By leveraging existing CADe systems, coordinates of regions or volumes of interest (ROI or VOI) are generated and function as input for a second tier, which is our focus in this study. In this second stage, we generate 2D (two-dimensional) or 2.5D views via sampling through scale transformations, random translations and rotations. These random views are used to train deep convolutional neural network (ConvNet) classifiers. In testing, the ConvNets assign class (e.g., lesion, pathology) probabilities for a new set of random views that are then averaged to compute a final per-candidate classification probability. This second tier behaves as a highly selective process to reject difficult false positives while preserving high sensitivities. The methods are evaluated on three data sets: 59 patients for sclerotic metastasis detection, 176 patients for lymph node detection, and 1,186 patients for colonic polyp detection. Experimental results show the ability of ConvNets to generalize well to different medical imaging CADe applications and scale elegantly to various data sets. Our proposed methods improve performance markedly in all cases. Sensitivities improved from 57% to 70%, 43% to 77%, and 58% to 75% at 3 FPs per patient for sclerotic metastases, lymph nodes and colonic polyps, respectively.

Sequential Monte Carlo tracking of the marginal artery by multiple cue fusion and random forest regression.

Given the potential importance of marginal artery localization in automated registration in computed tomography colonography (CTC), we have devised a semi-automated method of marginal vessel detection employing sequential Monte Carlo tracking (also known as particle filtering tracking) by multiple cue fusion based on intensity, vesselness, organ detection, and minimum spanning tree information for poorly enhanced vessel segments. We then employed a random forest algorithm for intelligent cue fusion and decision making which achieved high sensitivity and robustness. After applying a vessel pruning procedure to the tracking results, we achieved statistically significantly improved precision compared to a baseline Hessian detection method (2.7% versus 75.2%, p<0.001). This method also showed statistically significantly improved recall rate compared to a 2-cue baseline method using fewer vessel cues (30.7% versus 67.7%, p<0.001). These results demonstrate that marginal artery localization on CTC is feasible by combining a discriminative classifier (i.e., random forest) with a sequential Monte Carlo tracking mechanism. In so doing, we present the effective application of an anatomical probability map to vessel pruning as well as a supplementary spatial coordinate system for colonic segmentation and registration when this task has been confounded by colon lumen collapse.

A new 2.5D representation for lymph node detection using random sets of deep convolutional neural network observations.

Automated Lymph Node (LN) detection is an important clinical diagnostic task but very challenging due to the low contrast of surrounding structures in Computed Tomography (CT) and to their varying sizes, poses, shapes and sparsely distributed locations. State-of-the-art studies show the performance range of 52.9% sensitivity at 3.1 false-positives per volume (FP/vol.), or 60.9% at 6.1 FP/vol. for mediastinal LN, by one-shot boosting on 3D HAAR features. In this paper, we first operate a preliminary candidate generation stage, towards -100% sensitivity at the cost of high FP levels (-40 per patient), to harvest volumes of interest (VOI). Our 2.5D approach consequently decomposes any 3D VOI by resampling 2D reformatted orthogonal views N times, via scale, random translations, and rotations with respect to the VOI centroid coordinates. These random views are then used to train a deep Convolutional Neural Network (CNN) classifier. In testing, the CNN is employed to assign LN probabilities for all N random views that can be simply averaged (as a set) to compute the final classification probability per VOI. We validate the approach on two datasets: 90 CT volumes with 388 mediastinal LNs and 86 patients with 595 abdominal LNs. We achieve sensitivities of 70%/83% at 3 FP/vol. and 84%/90% at 6 FP/vol. in mediastinum and abdomen respectively, which drastically improves over the previous state-of-the-art work.

2D view aggregation for lymph node detection using a shallow hierarchy of linear classifiers.

Enlarged lymph nodes (LNs) can provide important information for cancer diagnosis, staging, and measuring treatment reactions, making automated detection a highly sought goal. In this paper, we propose a new algorithm representation of decomposing the LN detection problem into a set of 2D object detection subtasks on sampled CT slices, largely alleviating the curse of dimensionality issue. Our 2D detection can be effectively formulated as linear classification on a single image feature type of Histogram of Oriented Gradients (HOG), covering a moderate field-of-view of 45 by 45 voxels. We exploit both max-pooling and sparse linear fusion schemes to aggregate these 2D detection scores for the final 3D LN detection. In this manner, detection is more tractable and does not need to perform perfectly at instance level (as weak hypotheses) since our aggregation process will robustly harness collective information for LN detection. Two datasets (90 patients with 389 mediastinal LNs and 86 patients with 595 abdominal LNs) are used for validation. Cross-validation demonstrates 78.0% sensitivity at 6 false positives/volume (FP/vol.) (86.1% at 10 FP/vol.) and 73.1% sensitivity at 6 FP/vol. (87.2% at 10 FP/vol.), for the mediastinal and abdominal datasets respectively. Our results compare favorably to previous state-of-the-art methods.

Ligand-protein interactions of selective casein kinase 1δ inhibitors.

Casein kinase 1δ (CK1δ) and 1ε (CK1ε) are believed to be necessary enzymes for the regulation of circadian rhythms in all mammals. On the basis of our previously published work demonstrating a CK1ε-preferring compound to be an ineffective circadian clock modulator, we have synthesized a series of pyrazole-substitued pyridine inhibitors, selective for the CK1δ isoform. Additionally, using structure-based drug design, we have been able to exploit differences in the hinge region between CK1δ and p38 to find selective inhibitors that have minimal p38 activity. The SAR, brain exposure, and the effect of these inhibitors on mouse circadian rhythms are described. The in vivo evaluation of these inhibitors demonstrates that selective inhibition of CK1δ at sufficient central exposure levels is capable of modulating circadian rhythms.

Characterization of resilin-based materials for tissue engineering applications.

Modular proteins have emerged as powerful tools in tissue engineering because both the mechanical and biochemical properties can be precisely controlled through amino acid sequence. Resilin is an attractive candidate for use in modular proteins because it is well-known for having low stiffness, high fatigue lifetime, and high resilience. However, no studies have been conducted to assess resilin's compressive properties, cytocompatibility with clinically relevant cells, or effect on cell spreading. We designed a modular protein containing repeating sequences of a motif derived from Anopheles gambiae and cell-binding domains derived from fibronectin. Rapid cross-linking with tris(hydroxymethyl)phosphine was observed. The hydrogels had a complex modulus of 22 ± 1 kPa and yield strain of 63%. The elastic modulus in compression, or unconfined compressive modulus, was 2.4 ± 0.2 MPa, which is on the same order as human cartilage. A LIVE/DEAD assay demonstrated that human mesenchymal stem cells cultured on the resilin-based protein had a viability of 95% after three days. A cell-spreading assay revealed that the cells interacted with the fibronectin-derived domain in a sequence-specific manner and resulted in a mean cell area ~1.4-fold larger than when cells were seeded on a sequence-scrambled negative control protein. These results demonstrate that our resilin-based biomaterial is a promising biomaterial for cartilage tissue engineering.

Preventing rapid repeat births among latina adolescents: the role of parents.

Latina adolescent parents are at increased risk for rapid repeat births (second birth ≤ 24 months after the first), sexually transmitted infections, and negative educational and social outcomes. Although several effective parent-based interventions have been developed to prevent Latino youths' sexual risk taking, little research has explored the development of interventions to prevent repeat births that involve the parents of these adolescents. Existing preventative interventions involving parents suffer from important methodological limitations. Additional research is needed to advance theories of behavior, identify the causal pathways of parental influence, and specify appropriate behavioral targets. Future parent-based interventions to prevent repeat births should target pregnancy intentions, age of partners, contraceptive use, integrated prevention of pregnancies and sexually transmitted infections, educational attainment, and future orientations.

Modular cloning and protein expression of long, repetitive resilin-based proteins.

Resilin has emerged as a promising new biomaterial possessing attractive properties for tissue engineering applications. To date, proteins with repeating resilin motifs have been expressed with molecular weights less than 30 kDa. This work describes the development of resilin-based proteins (repeating motif derived from Anopheles gambiae) 50 kDa in size. A modular cloning scheme was utilized and features a recursive cloning technique that can seamlessly and precisely tune the number of resilin repeats. Previously-established resilin expression protocols (based on the Studier auto-induction method) were employed to express the proteins in Escherichia coli BL21(DE3)pLysS. Western blot and densitometry results demonstrated that only ~50% of expressed proteins were the desired molecular weight. This finding suggested that either protein truncation or degradation occurred during protein expression. Preventing leaky expression, lowering the culture temperature, and harvesting during exponential phase resulted in up to 94% of the expressed proteins having the desired molecular weight. These expression conditions differ from previously-published resilin expression methods and are recommended when expressing proteins with a larger number of repetitive resilin sequences.

Thiazole-diamides as potent gamma-secretase inhibitors.

The thiazole-diamide series (1) has been identified as highly potent gamma-secretase inhibitors. Several representative compounds showed IC(50) values of <0.3 nM. The synthesis and SAR, as well as a radiolabeled synthesis of [(3)H]-2a, are described.