skip to main content

This content will become publicly available on March 27, 2025

Title: Optimized Bags of Artificial Neural Networks Can Predict the Viability of Organisms Exposed to Nanoparticles
Prediction of organismal viability upon exposure to a nanoparticle in varying environments─as fully specified at the molecular scale─has emerged as a useful figure of merit in the design of engineered nanoparticles. We build on our earlier finding that a bag of artificial neural networks (ANNs) can provide such a prediction when such machines are trained with a relatively small data set (with ca. 200 examples). Therein, viabilities were predicted by consensus using the weighted means of the predictions from the bags. Here, we confirm the accuracy and precision of the prediction of nanoparticle viabilities using an optimized bag of ANNs over sets of data examples that had not previously been used in the training and validation process. We also introduce the viability strip, rather than a single value, as the prediction and construct it from the viability probability distribution of an ensemble of ANNs compatible with the data set. Specifically, the ensemble consists of the ANNs arising from subsets of the data set corresponding to different splittings between training and validation, and the different bags (k-folds). A k−1k machine uses a single partition (or bag) of k – 1 ANNs each trained on 1/k of the data to obtain a consensus prediction, and a k-bag machine quorum samples the k possible k−1k machines available for a given partition. We find that with increasing k in the k-bag or k−1k machines, the viability strips become more normally distributed and their predictions become more precise. Benchmark comparisons between ensembles of 4-bag machines and 34 fraction machines suggest that the 34 fraction machine has similar accuracy while overcoming some of the challenges arising from divergent ANNs in the 4-bag machines.  more » « less
Award ID(s):
Author(s) / Creator(s):
; ;
Publisher / Repository:
American Chemical Society
Date Published:
Journal Name:
The Journal of Physical Chemistry A
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. null (Ed.)
    We introduce a novel methodology for anomaly detection in time-series data. The method uses persistence diagrams and bottleneck distances to identify anomalies. Specifically, we generate multiple predictors by randomly bagging the data (reference bags), then for each data point replacing the data point for a randomly chosen point in each bag (modified bags). The predictors then are the set of bottleneck distances for the reference/modified bag pairs. We prove the stability of the predictors as the number of bags increases. We apply our methodology to traffic data and measure the performance for identifying known incidents. 
    more » « less
  2. Fadlelmola, Faisal Mohamed (Ed.)
    Sickle cell disease, a genetic disorder affecting a sizeable global demographic, manifests in sickle red blood cells (sRBCs) with altered shape and biomechanics. sRBCs show heightened adhesive interactions with inflamed endothelium, triggering painful vascular occlusion events. Numerous studies employ microfluidic-assay-based monitoring tools to quantify characteristics of adhered sRBCs from high resolution channel images. The current image analysis workflow relies on detailed morphological characterization and cell counting by a specially trained worker. This is time and labor intensive, and prone to user bias artifacts. Here we establish a morphology based classification scheme to identify two naturally arising sRBC subpopulations—deformable and non-deformable sRBCs—utilizing novel visual markers that link to underlying cell biomechanical properties and hold promise for clinically relevant insights. We then set up a standardized, reproducible, and fully automated image analysis workflow designed to carry out this classification. This relies on a two part deep neural network architecture that works in tandem for segmentation of channel images and classification of adhered cells into subtypes. Network training utilized an extensive data set of images generated by the SCD BioChip, a microfluidic assay which injects clinical whole blood samples into protein-functionalized microchannels, mimicking physiological conditions in the microvasculature. Here we carried out the assay with the sub-endothelial protein laminin. The machine learning approach segmented the resulting channel images with 99.1±0.3% mean IoU on the validation set across 5 k -folds, classified detected sRBCs with 96.0±0.3% mean accuracy on the validation set across 5 k -folds, and matched trained personnel in overall characterization of whole channel images with R 2 = 0.992, 0.987 and 0.834 for total, deformable and non-deformable sRBC counts respectively. Average analysis time per channel image was also improved by two orders of magnitude (∼ 2 minutes vs ∼ 2-3 hours) over manual characterization. Finally, the network results show an order of magnitude less variance in counts on repeat trials than humans. This kind of standardization is a prerequisite for the viability of any diagnostic technology, making our system suitable for affordable and high throughput disease monitoring. 
    more » « less
  3. null (Ed.)
    Abstract Inadequate at-home management and self-awareness of heart failure (HF) exacerbations are known to be leading causes of the greater than 1 million estimated HF-related hospitalizations in the USA alone. Most current at-home HF management protocols include paper guidelines or exploratory health applications that lack rigor and validation at the level of the individual patient. We report on a novel triage methodology that uses machine learning predictions for real-time detection and assessment of exacerbations. Medical specialist opinions on statistically and clinically comprehensive, simulated patient cases were used to train and validate prediction algorithms. Model performance was assessed by comparison to physician panel consensus in a representative, out-of-sample validation set of 100 vignettes. Algorithm prediction accuracy and safety indicators surpassed all individual specialists in identifying consensus opinion on existence/severity of exacerbations and appropriate treatment response. The algorithms also scored the highest sensitivity, specificity, and PPV when assessing the need for emergency care. Lay summary Here we develop a machine-learning approach for providing real-time decision support to adults diagnosed with congestive heart failure. The algorithm achieves higher exacerbation and triage classification performance than any individual physician when compared to physician consensus opinion. Graphical abstract 
    more » « less
  4. Multi-View Clustering (MVC) aims to find the cluster structure shared by multiple views of a particular dataset. Existing MVC methods mainly integrate the raw data from different views, while ignoring the high-level information. Thus, their performance may degrade due to the conflict between heterogeneous features and the noises existing in each individual view. To overcome this problem, we propose a novel Multi-View Ensemble Clustering (MVEC) framework to solve MVC in an Ensemble Clustering (EC) way, which generates Basic Partitions (BPs) for each view individually and seeks for a consensus partition among all the BPs. By this means, we naturally leverage the complementary information of multi-view data in the same partition space. Instead of directly fusing BPs, we employ the low-rank and sparse decomposition to explicitly consider the connection between different views and detect the noises in each view. Moreover, the spectral ensemble clustering task is also involved by our framework with a carefully designed constraint, making MVEC a unified optimization framework to achieve the final consensus partition. Experimental results on six real-world datasets show the efficacy of our approach compared with both MVC and EC methods.

    more » « less
  5. A major challenge in computational biology regards recognizing one or more biologically- active/native tertiary protein structures among thousands of physically-realistic structures generated via template-free protein structure prediction algorithms. Clustering structures based on structural similarity remains a popular approach. However, clustering orga- nizes structures into groups and does not directly provide a mechanism to select individual structures for prediction. In this paper, we provide a few algorithms for this selection prob- lem. We approach the problem under unsupervised multi-instance learning and address it in three stages, first organizing structures into bags, identifying relevant bags, and then drawing individual structures/instances from these bags. We present both non-parametric and parametric algorithms for drawing individual instances. In the latter, parameters are trained over training data and evaluated over testing data via rigorous metrics.

    more » « less