Genomic Selection Research Intern

@fabians73

Hello, I’ve read your comparative genomic-selection plan and I’m confident I can step in and keep the pipeline reproducible and moving. I have hands-on experience building end-to-end genomic prediction workflows in R: quality filtering and imputation (including PLINK and snp-ready R workflows), population-structure checks (PCA, kinship), and fitting/benchmarking rrBLUP, GBLUP, Random Forest, LightGBM, ElasticNet and CNN-style models adapted for SNP matrices. I’ll implement robust hyper-parameter grids, stratified cross-validation schemes, and clear predictive-ability metrics, then deliver well-commented R scripts, notebooks and comparative tables/figures in the shared Git repo. I’ll provide concise progress updates and short interpretive notes with each experiment batch to keep decisions evidence-driven. For next steps I can begin by auditing the current repo and running a baseline pipeline on a held-out subset within the first week. What are the target traits, current sample size and genotype format (VCF/PLINK/binary matrix), and do you have GPU access for CNN training? Thanks, Fabian

@Mikun007

Hello, Your genomic-selection project sounds very interesting. While my primary expertise is in Python-based data science and machine learning, I have strong experience working with data preprocessing, model benchmarking, and reproducible ML pipelines, which are directly relevant to this study. Relevant Experience -------------------------------- • Data preprocessing and cleaning (missing data handling, feature filtering, exploratory analysis) • Machine learning models such as Random Forest, Gradient Boosting (LightGBM/XGBoost), ElasticNet, and deep learning models • Designing cross-validation experiments and model comparison pipelines • Producing clear visualizations, performance metrics, and experiment summaries How I Can Contribute ---------------------------------- • Help design and manage the model benchmarking workflow • Run hyperparameter tuning and comparative evaluation of prediction models • Maintain clean, reproducible scripts and experiment tracking via Git • Provide concise analysis and visual reports of model performance I am comfortable quickly adapting to new research workflows and tools, and I can also pick up R-based scripts if needed while contributing to the modeling and analysis process. I’d be glad to learn more about your dataset and current pipeline.

@arvindk337

Your project on comparative genomic selection immediately caught my attention. Working with models like rrBLUP, GBLUP, Random Forest, LightGBM, CNN, and ElasticNet in a continuously expanding dataset is exactly the type of research-oriented pipeline I enjoy contributing to. I have experience building reproducible data science workflows in R, especially for experiments where datasets evolve over time and model benchmarking must remain consistent. For genomic-style pipelines, I typically structure the workflow into clear stages so results stay reproducible and easy to interpret