João Muzzi, PhD

Reconciliation of independent SF-1 perturbation studies

Conceived and executed an integrative strategy to reconcile three independent SF-1 perturbation studies that had reported highly divergent target-gene sets. I developed the threshold-optimization approach, introduced a methodological adjustment for a specific analytical limitation, performed the full analysis and interpretation, and led the manuscript writing.

The study used the NR5A1 regulon inferred in my 2022 regulatory-network work as an external cross-validation framework and demonstrated convergence at the transcriptional and phenotypic levels despite limited overlap among the originally reported genes.

The work was developed with the Institut de Pharmacologie Moléculaire et Cellulaire in France, received an Abstract Award at the International Adrenal Cancer Symposium at MD Anderson Cancer Center in 2024, and was published in the European Journal of Endocrinology in 2025.

SF-1 and β-catenin functional interaction

Lead the computational and transcriptional analysis of the functional interaction between SF-1 and β-catenin in adrenocortical carcinoma. The manuscript has been submitted and is currently under review.

Pediatric adrenocortical carcinoma cohort

Conduct exploratory clinical and molecular analyses of a pediatric ACC cohort. The study remains in progress, and results are not yet publicly disclosed.

Statistical collaboration across biomedical research

Provide statistical design, analysis, interpretation, and visualization for multidisciplinary projects from other research centers, contributing to studies in colorectal cancer, snakebite diagnostics, leprosy, and other biomedical areas.

Steroid-phenotype stratification and immune activity

Revisited a previously underused molecular classification and demonstrated that steroid phenotype is central to interpreting the immune landscape of ACC. Tumors with low steroidogenic activity showed stronger immune infiltration and activation markers, while highly steroidogenic tumors displayed a more immunosuppressive environment.

The work proposed steroid-phenotype stratification as a research framework for selecting patients more likely to benefit from immunotherapy and suggested that intratumoral steroidogenesis may need to be inhibited before immunotherapeutic approaches are considered in highly steroidogenic tumors.

It also identified a highly expressed immunotherapeutic target through pan-cancer comparison. Subsequent independent research supported the distinction between circulating cortisol levels and the intratumoral steroid environment, consistent with the biological hypothesis raised in the study.

Inference of the regulatory-network landscape of ACC

Reconstructed a transcriptional regulatory network for adrenocortical carcinoma and identified 369 regulons associated with overall survival. The analysis highlighted CENPA as a major proliferation-related regulator, ZBTB4 as a potential anti-tumorigenic and druggable regulator, and NR5A1/SF-1 as a central axis in ACC biology.

The NR5A1 regulon inferred in this study became a reusable analytical resource for subsequent work, including the 2025 cross-study reconciliation analysis and the ongoing SF-1/β-catenin investigation.

Breast-cancer lncRNA research

Contributed equally to an interdisciplinary study that identified immune-related long non-coding RNA signatures across breast-cancer molecular subtypes. The work expanded my research beyond adrenal tumors and has received more than 40 citations since publication in 2021.

Monitoramento Inteligente Terrestre

Transformed an early prototype into a production-ready MVP for monitoring structural changes in protected and public buildings through satellite imagery. I redesigned the architecture, improved and trained the model, built the interface, and deployed the complete solution.

Legal analytics and supplemental-fee management

Built a production legal analytics platform used by the Brazilian Bar Association — Paraná Section. The system automated the collection and organization of public procedural information, used AI-assisted document-access workflows, structured the relational database, and implemented the logic required for investigation, classification, monitoring, and operational use.

The platform processed more than one million judicial records and supported the investigation and classification of hundreds of thousands of proceedings involving tens of thousands of attorneys.

Institutional and public-finance analytics

Developed simulation and impact-assessment models related to RPV thresholds, court-ordered debt, state revenue, and other institutional financial questions, supporting technical interpretation, negotiation, and strategic decision-making.

TOTVS/Protheus and VCOM technical consulting

Provide technical consulting to OAB-PR on the implementation of VCOM and the stabilization and integration of TOTVS Protheus workflows.

AI-assisted Genetic-risk platform

Designed and implemented a workflow that processed VCF files, selected and annotated clinically relevant variants, interpreted disease-associated evidence, and generated individualized reports. Built a literature-based genetic-risk framework using curated variants and published evidence. Developed a neural-network model after selecting the most informative variants for risk classification. The workflow is progressing toward experimental validation.

Institutional revenue forecasting

Developed a SARIMAX monthly time-series model using historical data from 2010 to 2023. The monthly projections were aggregated into annual estimates and achieved approximately 97% agreement with realized annual revenue in the validation period, outperforming the linear, polynomial, and other regression-based alternatives evaluated.

Judicial-fee decision support

Led the quantitative modeling of alternative judicial-fee structures in a project involving the Brazilian Bar Association — Paraná Section and Paraná’s justice system. I translated legal and financial proposals into measurable scenarios and developed an interactive simulation comparing percentage rates, minimum charges, and maximum caps.

The analysis supported technical discussions and contributed to the development of a proposal subsequently approved by the Legislative Assembly of Paraná.

Precatório simulation and PEC 66 analysis

Developed a public Shiny simulator requested by OAB-PR to demonstrate how payment limits, debt correction, revenue growth, and new precatório inflows could affect state and municipal debt over time.

Corporate due diligence platform

Designed and coordinated the company’s internal investigation platform, integrating public information from Brazil’s Federal Revenue Service and PGFN to support analysis of corporate relationships, tax liabilities, economic groups, and business backgrounds.

The solution processed approximately 66 million public-data records and involved ETL, relational database architecture, entity linkage, validation, automation, and user-facing analytical workflows.

Complex financial-investigation data processing

Designed the technical process for extracting, ordering, linking, and interpreting large volumes of financial and corporate data in a complex forensic investigation.

OAB-PR analytics and systems consulting

Delivered recurring reports, management-impact studies, financial analyses, business intelligence, and an API for automated extraction from TOTVS Protheus. I also coordinated internal automations and the company intranet with support from two data-science interns.

Regulatory accountability report

Delivered an interactive regulatory report within one week to support an independent verifier’s statutory reporting requirements. It analyzed more than 200,000 tons of reverse-logistics credits documented through invoices, including origin, waste type, tonnage, major generators, duplicate records, and potentially invalid invoices.

Arcos Dorados oil-collection analytics

Developed a dashboard relating unit revenue to used-oil generation, identified three behavioral clusters, and flagged units with unexpectedly low collection volume relative to sales. Historical monitoring and email alerts were followed by an increase of more than 15% in collected oil volume.

State-system integration monitoring

Created a daily monitoring system for integrations between the company platform and state environmental systems. It measured success rates by state and delivered failed MTRs, client details, error messages, and logs to technical and support teams.

The workflow replaced delayed client complaints with proactive daily detection. Integration success improved from approximately 50% in the first week of monitoring to more than 97.5% within a one-week operational window by the time I left the company.

Automated reverse-logistics credit validation

Coordinated one of Brazil’s first automated workflows for validating reverse-logistics credits, including invoice verification, duplicate detection, data consistency, and accountability reporting.

Led a project team of four developers and coordinated the required design activities.

Database restructuring

Analyzed a complex relational database containing hundreds of tables and partially documented relationships. Coordinated restructuring and query refactoring across one-to-one, one-to-many, many-to-one, and many-to-many structures.

SABESP privatization market study

Produced a technical market and lead analysis used by a client to evaluate whether to pursue an acquisition opportunity related to the privatization of SABESP and to estimate an appropriate investment range.

FleetGreen MVP

Independently developed the technical MVP of FleetGreen in approximately four months in collaboration with Mercedes-Benz. I created the API integration with the telemetry platform, defined emissions indicators and analytical logic, processed the data, and built the original dashboards and product functionality.

A redesigned presentation layer was later prepared for FENATRAN 2022 while retaining the core functionality, metrics, indicators, and concepts established in the MVP.

FleetGreen subsequently evolved under Mercedes-Benz and received third place in the Green Products category at the 2024 Daimler Truck Green Mover Award.

End-to-end operational understanding

Worked across operational, commercial, administrative, and financial activities, developing a practical understanding of the business before leading its process digitalization and analytical restructuring.

Quotation and pricing automation

Replaced paper-based quotations with a standardized spreadsheet system connected to price tables. Users entered measurements and component codes, and the model calculated dimensions, materials, service values, and the final client-facing layout automatically.

The system substantially accelerated preparation and revision of up to approximately 200 monthly quotations, reduced manual errors, and made updates nearly immediate after supplier price changes.

Cost and freight modeling

Developed pricing formulas combining direct material costs with labor, infrastructure, and variable expenses, as well as freight-pricing models based on distance, load, and operating cost.

ABC analysis and strategic concentration risk

Conducted ABC analyses of expenses and clients and identified a significant concentration of revenue among large construction companies. The analysis highlighted a strategic vulnerability that became increasingly relevant during the subsequent downturn in the construction sector.

Business diversification

Helped create new revenue streams, including aluminum fabrication and frameless glass systems. Aluminum fabrication became particularly important to sustaining the company during its later years.

Environmental-control stabilization

Investigated a long-standing environmental-control problem that had remained unresolved for several years. I inspected physical sensors, mapped sensor-to-software relationships, recovered previously unused historical reports, and identified limitations in the original installation.

Because the system could not fully automate simultaneous control of humidity, temperature, and differential pressure, I developed a structured manual adjustment process. This restored partial operational stability and kept key variables within required production ranges despite infrastructure limitations.

UNESCO BIODEV research project

Completed the academic credits of the UNESCO BIODEV Master of Science program and developed a project for a quantitative tetanus-toxoid detection method. The planned R&D structure was not implemented, and the Bacterial Vaccines Division was later discontinued, preventing completion of the experimental phase.

Fermentation set-point modeling

Modeled aeration and nutrient-dosage parameters, implemented new set points, and monitored production data to confirm that the automated process behaved as expected under real operating conditions.

Plant-wide electricity allocation model

Developed a model combining direct equipment measurements with operating time, installed power, indirect indicators, and product-level production variables to estimate missing energy consumption and allocate electricity costs across products.

The model predicted the following month’s total electricity bill with 99.5% agreement and was subsequently used to support bill-of-material calculations, production-cost comparisons among international plants, and evaluation of Brazilian energy costs in manufacturing decisions.

Laboratory setup and training

Joined as the laboratory’s first undergraduate student and, under faculty supervision, helped standardize procedures, document routines, and train approximately ten incoming undergraduate and graduate students in methods they had not previously used.

Differential immunodiagnostics

Worked on the development of a diagnostic approach to distinguish neurocysticercosis from toxoplasmosis using immunochemical and serological methods.

Phage-display collaboration

Contributed experimental ELISA work to a phage-display project involving bovine sera, which later resulted in a peer-reviewed publication on peptide-based detection of Taenia saginata infection.