Jason D. Wittenbach

ORCID: 0000-0003-2937-9758
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Neurobiology and Insect Physiology Research
  • Animal Vocal Communication and Behavior
  • Stock Market Forecasting Methods
  • Animal Behavior and Reproduction
  • Time Series Analysis and Forecasting
  • Marine animal studies overview
  • Explainable Artificial Intelligence (XAI)
  • Neural dynamics and brain function
  • Physiological and biochemical adaptations
  • Insect and Arachnid Ecology and Behavior
  • Neuroscience and Neural Engineering
  • Complex Systems and Time Series Analysis
  • Anomaly Detection Techniques and Applications
  • Machine Learning and Data Classification
  • Stroke Rehabilitation and Recovery
  • Zebrafish Biomedical Research Applications
  • Actinomycetales infections and treatment
  • Advanced Statistical Methods and Models
  • Astronomical and nuclear sciences
  • Invertebrate Immune Response Mechanisms
  • interferon and immune responses
  • Data Stream Mining Techniques
  • Adversarial Robustness in Machine Learning
  • Meta-analysis and systematic reviews
  • Single-cell and spatial transcriptomics

Zero to Three
2023

Janelia Research Campus
2018-2022

Howard Hughes Medical Institute
2018-2022

University of California, Irvine
2019

Cure CMD
2019

Congenital Muscle Disease International Registry
2019

Alberta Children's Hospital
2019

University of Rochester
2019

St Mary's University College
2019

University of Cambridge
2019

Abstract Molecular profiles of neurons influence neural development and function but bridging the gap between genes, circuits, behavior has been very difficult. Here we used single cell RNAseq to generate a complete gene expression atlas Drosophila larval central nervous system composed 131,077 cells across three developmental stages (1 h, 24 h 48 after hatching). We identify 67 distinct clusters based on patterns expression. These include 31 functional mature neuron clusters, 1 ring gland...

10.1186/s13064-022-00164-6 article EN cc-by Neural Development 2022-08-24

Variable motor sequences of animals are often structured and can be described by probabilistic transition rules between action elements. Examples include the songs many songbird species such as Bengalese finch, which consist stereotypical syllables sequenced according to (song syntax). The neural mechanisms behind poorly understood. Here, we investigate where song syntax is encoded in brain finch rapidly reversibly manipulating temperature production pathway. Cooling premotor nucleus HVC...

10.1523/jneurosci.1827-16.2017 article EN cc-by-nc-sa Journal of Neuroscience 2017-02-03

Abstract Background Early pathogen ID and targeted treatment are key to reducing bloodstream infection (BSI) morbidity mortality. Current diagnostics rely on culture which takes 1-2 days for longer antimicrobial susceptibility testing (AST) results, or molecular assays with limited panels, few resistance markers, high false positive rates. We report interim results of a first-in-kind comprehensive predictive AST assay directly from patient blood samples. The system can deliver in ∼8 hrs,...

10.1093/ofid/ofae631.315 article EN cc-by Open Forum Infectious Diseases 2025-01-29

Consecutive repetition of actions is common in behavioral sequences. Although integration sensory feedback with internal motor programs important for sequence generation, if and how contributes to repetitive poorly understood. Here we study auditory generating syllable sequences songbirds. We propose that signals provide positive ongoing commands, but this influence decays as weakens from response adaptation during repetitions. Computational models show mechanism explains repeat...

10.1371/journal.pcbi.1004471 article EN cc-by PLoS Computational Biology 2015-10-08

To better characterize adult myotubularin 1 (MTM1)-related myopathy carriers and recommend a phenotypic classification.This cohort study was performed at the NIH Clinical Center. Participants were required to carry confirmed MTM1 mutation recruited via Congenital Muscle Disease International Registry (n = 8), traveling local clinic of Neuromuscular Neurogenetic Disorders Childhood Section, National Institute Neurological Stroke, Cure CMD 1), direct physician referral 1). examinations, muscle...

10.1212/wnl.0000000000008316 article EN Neurology 2019-09-21

Abstract Large, multivariate geographic datasets have been used to characterize space with the help of spatial data mining tools. In our study, we explore sufficiency S upport V ector M achine ( SVM ), a popular machine‐learning technique for unsupervised classification and clustering, recognize hidden patterns in college admissions dataset. Our dataset holds over 10,000 students applying an undisclosed university during one year. Students are qualified almost exclusively by their...

10.1111/j.1467-9671.2012.01354.x article EN Transactions in GIS 2012-10-09

Summary Molecular profiles of neurons influence information processing, but bridging the gap between genes, circuits, and behavior has been very difficult. Furthermore, behavioral state an animal continuously changes across development as a result sensory experience. How influences molecular cell is poorly understood. Here we present complete atlas Drosophila larval central nervous system composed over 200,000 single cells four developmental stages. We develop polyseq , python package, to...

10.1101/785931 preprint EN cc-by-nc-nd bioRxiv (Cold Spring Harbor Laboratory) 2019-09-30

In the environment of fair lending laws and General Data Protection Regulation (GDPR), ability to explain a model's prediction is paramount importance. High quality explanations are first step in assessing fairness. Counterfactuals valuable tools for explainability. They provide actionable, comprehensible individual who subject decisions made from prediction. It important find baseline producing them. We propose simple method generating counterfactuals by using gradient descent search latent...

10.48550/arxiv.2012.09301 preprint EN cc-by arXiv (Cornell University) 2020-01-01

Muscle strength testing is routine in clinical practice. Here we provide an aid to the documentation and visual conceptualization of those results - MuscleViz: a free, open-source application for visualizing muscle testing. Its use settings streamlines communication physical examination findings. The tool also useful presenting patient data case reports or series. A push towards software has benefitted other areas science; believe similar effort dedicated development tools worth pursuing.

10.3233/jnd-190385 article EN Journal of Neuromuscular Diseases 2019-05-21

In data science, there is a long history of using synthetic for method development, feature selection and engineering. Our current interest in comes from recent work explainability. Today's datasets are typically larger more complex - requiring less interpretable models. the setting \textit{post hoc} explainability, no ground truth explanations. Inspired by explaining image classifiers that does provide truth, we propose similar solution tabular data. Using copulas, concise specification...

10.48550/arxiv.2007.10532 preprint EN other-oa arXiv (Cornell University) 2020-01-01

Abstract Background Whole genome sequencing (WGS) of bacterial pathogens directly from clinical samples or cultured isolates is a promising technique for diagnosing infections, yet challenges remain in using WGS to guide antimicrobial therapy. The standard approach predicts susceptibility test (AST) profiles the presence/absence known resistance markers. Here we measure performance markers compared machine learning method predicting AST. Methods We assessed ResFinder, publicly available...

10.1093/ofid/ofad500.2461 article EN cc-by Open Forum Infectious Diseases 2023-11-27

Behavioral sequences of animals are often structured and can be described by probabilistic rules (or "action syntax"). The patterns vocal elements in birdsong a prime example. encoding such neural circuits is poorly understood. Here we locate the site song syntax Bengalese finch rapidly reversibly manipulating temperature production pathway. Changing premotor nucleus HVC (proper name) alters transition probabilities between syllables. Most prominently, cooling reduces number repetitions long...

10.48550/arxiv.1501.06108 preprint EN other-oa arXiv (Cornell University) 2015-01-01

The majority of cortical synapses are local and excitatory. Local recurrent circuits could implement amplification, allowing for pattern completion other computations 1 . Cortical contain subnetworks, consisting neurons with similar receptive fields elevated connectivity relative to the network average 2,3 Understanding performed by these subnetworks during behavior has been hampered fact that encoding different types information spatially intermingled distributed over large brain volumes...

10.1101/822700 preprint EN cc-by-nc-nd bioRxiv (Cold Spring Harbor Laboratory) 2019-10-29

Temporal data are ubiquitous in the financial services (FS) industry -- traditional like economic indicators, operational such as bank account transactions, and modern sources website clickstreams all of these occur a time-indexed sequence. But machine learning efforts FS often fail to for temporal richness data, even cases where domain knowledge suggests that precise patterns between events should contain valuable information. At best, treated uniform time series, there is sequence but no...

10.48550/arxiv.2009.05636 preprint EN other-oa arXiv (Cornell University) 2020-01-01
Coming Soon ...