Marcelo Mollinari, PhD - Consultant, Quantitative Genetics & Bioinformatics

About Me

I am a consultant in quantitative genetics and bioinformatics at AbacusBio, helping breeding programs turn complex data into reliable selection decisions. My work focuses on genetic evaluation, genomic prediction, and practical analytics that improve accuracy, shorten breeding cycles, and support cost-effective deployment strategies.

I design and implement reproducible workflows for genotype and phenotype curation, mixed-model analysis, model monitoring, and decision-ready reporting. Across diploid and polyploid contexts, I combine statistical rigor with production-ready R/C++ tooling so teams can move from analysis to action.

Capabilities

Genetic evaluation: BLUP/GBLUP pipelines, multi-trait and multi-environment models, spatial/heterogeneity effects, and model diagnostics for operational decisions.
Genomic prediction: Cross-validation design, training population design, and ongoing monitoring of prediction accuracy and model drift.
Breeding decision support: Selection indices, economic weighting, genotyping strategy, and cost-benefit tradeoffs aligned with program goals.
Polyploid analytics: Robust handling of dosage uncertainty, mapping/phasing pitfalls, and interpretation risks in complex ploidy systems.
Software & reproducibility: Production-focused R/C++ tooling, package development, automated pipelines, and reporting for repeatable analysis workflows.

Workflow of MAPpoly2 and GGSpoly — Workflow illustrating the implementation of the models and tools developed in MAPpoly2 and GGSpoly.

Selected Work

Built genomic prediction workflows for multi-environment breeding trials, including validation and deployment checkpoints.
Developed genotype data QC and curation pipelines with automated flags and reproducible reporting.
Implemented mixed-model evaluation and reporting for routine selection decisions in operational breeding contexts.
Designed decision-support analyses for trait weighting, candidate ranking, and genotyping investment planning.

Representative examples. Details available upon request.

Open-source software

These open-source projects are independent of client engagements and shared for research and community use.

These are some software tools I have maintained and contributed to:

– SIMpoly - Simulation of Genetic Marker Data in Mixed-Ploidy Level Pedigreed Populations GitHub Repository (creator)

SIMpoly This is a new addition to my package suite - a simple tool designed to simulate interconnected populations for testing mappoly2. I have no plans to publish or submit it to CRAN, but I’m making it available in case others find it useful.
– Genetic maps in interconnected diploid and autopolyploid families. GitHub Repository (creator)

MAPpoly2 is an advanced R package designed for constructing genetic maps in interconnected full-sib autopolyploid families. This enhanced version prioritizes user-friendliness and accessibility, with improvements geared toward seamless integration with tools like R Shiny to provide an intuitive interface for polyploid genetic mapping. The package supports ploidy levels of 2, 4, and 6, allowing for any combination of these configurations, making it highly versatile for diverse research needs.

A standout feature of MAPpoly2 is its significantly enhanced performance, achieved through the integration of computationally intensive code implemented in C++. This optimization enables the efficient analysis of large and complex datasets, addressing a common bottleneck in genetic mapping workflows.

The package employs a robust Hidden Markov Model (HMM) framework to construct individual genetic maps for each parent. These maps are then combined into a comprehensive joint map, which is recomputed to incorporate any remaining markers, ensuring high accuracy and resolution in the final results. These innovations make MAPpoly2 an invaluable tool for researchers working with polyploid species and large-scale genomic datasets.
– Genetic maps in diploids and autopolyploids. CRAN Package (creator)

MAPpoly is a tool for genetic mapping in diploids and autopolyploids, providing statistical models...
– Visualization of Genetic Analysis in Polyploids. CRAN Package (conceptual lead)

VIEWpoly offers a powerful graphical user interface designed for integrating, visualizing, and exploring results from linkage and quantitative trait loci (QTL) analyses, along with genomic data, specifically tailored for autopolyploid and diploid species. This interactive application allows users to seamlessly upload and combine various data sources, including gene annotations and alignment files, to facilitate candidate gene exploration in an integrated genome browser.

VIEWpoly supports input from multiple tools beyond MAPpoly, including polymapR, diaQTL, QTLpoly, and polyqtlR, making it a versatile solution for researchers working with complex polyploid genetics. By enabling the integration of diverse datasets, VIEWpoly provides an interactive platform for comprehensive analysis and discovery in polyploid research.
– Random-Effect Multiple QTL Mapping in Autopolyploids. CRAN Package (contributor)

QTLpoly The R package QTLpoly is software to map quantitative trait loci (QTL) in full-sib families of outcrossing autopolyploid species based on a random-effect multiple QTL model (Pereira et al. 2020). Variance components associated with putative QTL are tested using score statistics (Qu et al. 2013), and final models are fitted using residual maximum likelihood (REML, adapted from the R package sommer).
– Genetic mappings in experimental crosses. CRAN Package (co-creator)

OneMap is a software package designed for constructing genetic maps in experimental crosses, including full-sib, RILs, F2, and backcross populations. Originally developed by Gabriel R. A. Margarido, myself (Marcelo Mollinari), and A. Augusto F. Garcia, the project later welcomed contributions from Rodrigo R. Amadeu, Cristiane H. Taniguti, and Getulio C. Ferreira. OneMap has been available on CRAN for several years, with its most recent stable version released in 2020. The CRAN version is recommended for most users, while the development version—featuring ongoing improvements and experimental functions—is hosted on GitHub. A major milestone for OneMap was the integration of support for sequencing-based markers, such as those from Illumina and GBS technologies, allowing users to analyze modern genomic datasets effectively. The package provides a comprehensive set of functions for building linkage maps, with some specialized functions running internally to optimize analysis workflows.
– Web-based platform for genomic data analysis in diploid and polyploid species. GitHub Repository (co-creator)

GGSpoly is a web-based tool for genomic data analysis in diploid and polyploid species. It provides an intuitive interface for breeders and researchers to upload, analyze, and visualize genomic data. The platform supports various genetic analysis tasks, including haplotype inference, QTL mapping, and consensus map construction. GGSpoly aims to bridge the gap between complex genetic models and practical breeding applications, making advanced genomic tools accessible to a broader audience.

Selected Publications

Mollinari M, Garcia AAF. (2019). Linkage analysis and haplotype phasing in experimental autopolyploid populations with high ploidy level using hidden Markov models. G3: Genes, Genomes, Genetics, https://doi.org/10.1534/g3.119.400378.
Mollinari M, Olukolu BA, Pereira G da S, Khan A, Gemenet DC, Yencho GC, Zeng ZB. (2020). Unraveling the hexaploid sweetpotato inheritance using ultra-dense multilocus mapping. G3: Genes, Genomes, Genetics, https://doi.org/10.1534/g3.119.400620.
Taniguti CH, Gesteira G de S, Lau J, Pereira G da S, Zeng Z-B, Byrne DH, Riera-Lizarazu O, Mollinari M. (2023). VIEWpoly: a visualization tool to integrate and explore results of polyploid genetic analysis. J. Open Source Softw., 7(74), 4242. https://doi.org/10.21105/joss.04242.
Wu S, Sun H, Hamilton JP, Mollinari M, Gesteira SG, Kitavi M, Yan M, Kreuze J, Yencho GC, Buell R, Fei Z. (2024). Phased chromosome-level genome assembly offers new insights into the origin of hexaploid sweetpotato. bioRxiv. https://doi.org/10.1101/2024.08.17.608395. Under review.
Gesteira GS, Pereira GS, Zeng ZB, Mollinari M. (2024). Genetic maps in sweetpotato. In The Sweetpotato Genome (pp. 45-68). https://doi.org/10.1007/978-3-031-65003-1_5.

See Google Scholar for complete list of publications

Talks, Workshops and Teaching

Here are slides from some of the talks, workshops, and lectures I've given over the years:

Theoretical and Practical Aspects of Linkage Analysis – A workshop presentation from the Breeding Insight Workshop at Cornell covering linkage analysis in diploids and polyploids, including genetic segregation, genotype calling, and polyploid inheritance patterns.
Genomic Challenges in Polyploid Crops – Overview of the difficulties in genomic analysis and breeding of polyploids, covering variant discovery, genotype calling, and QTL mapping strategies.
HS 703: Importance of Polyploidy, Inheritance Patterns, and Manipulation – Lecture discussing fundamental aspects of polyploid genetics, including meiotic pairing, gamete formation, and methods for analyzing segregation in polyploids.
Introduction to MAPpoly – A training session on the MAPpoly software for genetic mapping in polyploids, detailing installation, data import, filtering, and map construction.
Merging Maps from Multiple Populations – Explores methods for integrating genetic maps from multiple populations, with emphasis on recombination fractions, linkage phases, and marker ordering.
The Sweetpotato Genome – Our book discusses the sequencing efforts and genetic research on sweetpotato, including its significance for food security, genomic resources, and marker-assisted breeding.
Genetic Mapping in Polyploids: From Genotyping to Haplotype Reconstruction – Lecture at Universidade Federal de Goiás explaining key concepts in genetic mapping of polyploids, including linkage analysis, meiotic configurations, and computational tools.
Updates: Introducing MAPpoly2 and QTLpoly – A presentation on recent developments in MAPpoly2 and QTLpoly, focusing on mixed-ploidy crosses, haplotype probabilities, and new computational methods.

Miscellaneous Scripts

Explore additional scripts and resources related to my research and software development:

Practical guide to integrating C++ with R using Rcpp – This may be a decade old, but I still find it valuable—a practical guide to integrating C++ with R using Rcpp, complete with examples, performance comparisons, and package development insights.
Miscellaneous Scripts on GitHub – A collection of scripts and resources for various bioinformatics and genetic analysis tasks.

Contact

For consulting inquiries in quantitative genetics and bioinformatics, please get in touch: