Analytics, Biostatistics, and Machine Learning (ABML) module

Services

To receive services from this module, contact Marlene Behrmann, Alireza Chamanzar, or Pirro Hysi.

Office Hours: 

• Pirro Hysi: Most Wednesdays 4 - 5 pm - Office 7.387
• Alireza Chamanzar: Fridays 12 – 1 pm -Office 8.331B
• For specifics, including updates and cancellations, please check the ABML Office Hours shared calendar in Outlook.

Module Description 

The ABML module is the newest addition to P30 core offerings, developed after a faculty survey overwhelmingly indicated a need for analytical and computational support. Our faculty generate a tremendous amount of data from a variety of sources, including de-identified medical records, imaging (retinal, tissue, brain MRI and connectivity imaging), neuro- and electrophysiological recordings from members of our cortical vision faculty, performance-based measures, molecular data from metabolomic, microbiome, RNA-Seq, ATAC-Seq, Single-cell/nuclei RNA- and ATAC-Seq, and ChIP-Seq. There is a great need for additional support to fully utilize this data to its highest potential by leveraging departmental expertise in cutting-edge statistical methods and tools.
 
The ABML module will connect people generating data with module personnel who are experts in biostatistics, statistical modeling, machine learning, and deep learning methodologies. Drs. Pirro Hysi and Alireza Chamanzar will be available to advise on study design, statistical options, analytical choices and to provide guidance on hands-on help available to all members of the department. Both of these experts will offer regular office hours and are also available to schedule appointments. Additionally, the ABML module will offer regular lunch-and-learns and seminars throughout the year to provide more details on special topics and to crowdsource solutions to specific data analytics challenges.

More information on the AMBL consultants and their areas of expertise can be found below.

Alireza Chamanzar

The Chamanzar Lab provides high-impact data analytics and data science consulting grounded in a rigorously data-driven approach. Our expertise spans advanced data analytics, machine learning, statistical analysis, and signal processing to transform large, complex, and multimodal datasets—including high-dimensional time-series (e.g., scalp and intracranial EEG, fNIRS, fMRI, etc.) and imaging data (e.g., OCT, WSI, and radiology images)—into actionable insights. 
We deliver end-to-end, data-driven solutions covering data curation, feature engineering, dimensionality reduction, predictive modeling, and rigorous validation, with a strong emphasis on robustness, reproducibility, and interpretability. By combining deep domain knowledge with data-centric methodologies, we help our colleagues at Vision Institute to develop scalable, evidence-based strategies that translate real-world data into measurable scientific, clinical, and operational impact.

Keywords: signal processing, statistical machine learning (ML), and non-generative deep learning (DL)

Pirro Hysi

Dr. Hysi has a research interest in biological biomarker analyses, particularly genetics, transcriptomics, and metabolomics, as well as models incorporating information from multiple high throughput ‘omics’ platforms.

He has considerable expertise in applied statistics and statistical (machine) learning models. He will be available to advise on various aspects of your projects, from early methodological guidance, study design, sample ascertainment, data collection strategies, choice of appropriate statistical models, statistical power projections and planning, quality assurance for the analyses, to help with interpretation of results.

Please contact Dr. Hysi with any queries or questions via email or simply drop in for a conversation at any of the times shown on the calendar.

Lingji Zhu

Lingji Zhu is a Research Associate working with Drs. Chamanzar and Hysi to provide hands-on statistical analysis of data provided by researchers in the Department of Ophthalmology. She is skilled at R and Python, as well as data pre-processing activities. Lingji’s participation ensures that the ABML module is able to offer not only high-level consultation but also direct analysis of specific datasets to expand the capabilities of research undertaken at the Vision Institute.