Colin Shepherd - Portfolio & CV

Skills

Data Science & AI

R & Python (Bioconductor packages, Pandas, NumPy...)
Machine learning algorithms (caret, scikit-learn, PyTorch)
Natural language processing (SpaCy, Huggingface)
Data visualization & interpretation (ggplot2, matplotlib)
Statistical analysis
Algorithms

Tools & Others

Git, GitHub/GitLab
Docker
PyTorch and scikit-learn
Firebase
Google Cloud
Jupyter

App Development

Dart, Flutter SDK
State Management
Firebase Integration
Native Device Features
Security, scalability and reliability

About Me

I'm a dedicated and innovative problem solver and can help you to exploit your data to inform your decision making.

I have a PhD in Molecular and cellular biology and many years of experience in scientific research. As well as the biology, I can program, handle and analyse data appropriately, use AI and machine learning to enhance analyses.

I thrive in environments where I can apply my skills in data analysis and machine learning to create intelligent, user-centric solutions. This portfolio showcases my work, demonstrating my capabilities and commitment to delivering solutions.

Please explore my page and feel free to reach out if you'd like to collaborate or discuss opportunities.

Work Experience

June 2025 - Present

Director at AdaptiveML

Remote

Developed real-time data analysis pipelines for collecting, processing and analysing data.
Deployed cloud-based machine learning models to production environments.
Implemented real-time predictions and automated decision-making systems.
Led the development and delivery of intelligent data-driven solutions.

Jan 2019 - August 2025

Scientific Editor at Fios Genomics

Edinburgh, UK

Worked across genomics and transcriptomics (and other -omics) technologies and data sets.
Applied appropriate statistical analyses and quality evaluation to various data sets.
Trained and applied machine learning models to real data sets.
Programming in R and Python.
Interpretation and discussion of results in context and how to use these to direct future work.

Sept 2013 - Jan 2019

Postdoctoral researcher at Newcastle University

Newcastle-upon-Tyne, UK

Worked on several projects which investigated the links between genetic variation in DNA, disease and treatments.
First appreciated the power of computer programming.

About My Work at AdaptiveML

I develop intelligent data-driven solutions that transform raw data into actionable insights. Here are the key projects I'm working on:

Real-Time Charting and ML Prediction

In Production

Hourly BTC/USD forecast demonstrating real-time ML deployment and data visualisation

NLP-Driven Sales Forecasting

Completed Q3 2025

Advanced NLP and ML pipeline for sales forecasting and strategic recruitment with real-world deployment

Market News Sentiment Analysis

Active Q4 2025

Real-time sentiment analysis of financial news using BERT transformer models tracking S&P500, NASDAQ, DOW, FTSE100 and GOLD

Equipment Health Monitor

Active

Real-time equipment health monitoring with live vibration data from ESP32 sensors, cloud-based aggregation and automated maintenance tracking

TCGA BRCA Data Analysis

Completed 2025

ML and deep learning analysis of TCGA-BRCA RNA-Seq dataset using KNN and deep learning models for breast cancer subtype classification

View case study

Education

PhD Molecular and Cellular Biology

University of Glasgow | UK

Graduated: July 2014

BSc (Hons) Biochemistry

University of Glasgow | UK

Graduated: July 2010

Certificates

Finding Hidden Messages in DNA (Bioinformatics I)

A series of classes illustrating the power of computing in modern biology. Essentially it was an advanced R programming course where I learned to use scripting and automated processes to solve complex biological data problems.

Advanced Data Visualisation with R

Coursera certification. Included advanced figures with ggplot2, spatial data, plotly and gganimate.

R Programming

Coursera certification. I learned general programming and scripting with R, loop functions and debugging, simulation and profiling, and R markdown.

The Data Scientist’s Toolbox

Coursera certification. Here modules included data science fundamentals, R and RStudio, version control and Github, R markdown and big data.

Prioritization of PLEC and GRINA as Osteoarthritis Risk Genes Through the Identification and Characterization of Novel Methylation Quantitative Trait Loci

Authors: SJ Rice, M Tselepi, AK Sorial, G Aubourg, C Shepherd, D Almarza, et al.

Journal: Arthritis & Rheumatology, 2019

Expression analysis of the osteoarthritis genetic susceptibility mapping to the matrix Gla protein gene MGP

Authors: C Shepherd, AE Reese, LN Reynard, J Loughlin

Journal: Arthritis Research & Therapy, 2019

Prioritization of PLEC and GRINA as osteoarthritis risk genes through the identification and characterization of novel methylation quantitative trait loci 2019

Authors: S Rice, G Aubourg, T Sorial, DA Gomez, M Tselepi, C Shepherd, et al.

Type: Report/Conference Abstract, 2019

Multicentre genome wide association study identifies risk alleles for progressive chronic lymphocytic leukaemia

Authors: D Allsup, WY Lin, SE Fordham, N Sunter, C Elstob, T Rahman, E Willmore, et al.

Journal: Blood, 2019

Functional Characterization of the Osteoarthritis Genetic Risk Residing at ALDH1A2 Identifies rs12915901 as a Key Target Variant

Authors: C Shepherd, D Zhu, AJ Skelton, J Combe, H Threadgold, L Zhu, et al.

Journal: Arthritis & Rheumatology, 2018

Genotype at the MGP OA risk locus correlates with an expression quantitative trait locus operating on MGP in cartilage and with dna methylation at a cluster of regulatory CpGs

Authors: C Shepherd, LN Reynard, J Loughlin

Journal: Osteoarthritis and Cartilage, 2018

Identification of novel methylation quantitative trait loci (mqtls) and functional characterization using CRISPR/CAS9 and gene expression analysis prioritizes Plec as an oa gene

Authors: SJ Rice, AK Sorial, G Aubourg, C Shepherd, M Tselepi, D Almarza, et al.

Journal: Osteoarthritis and Cartilage, 2018

Functional characterisation of the osteoarthritis genetic risk locus that resides at the gene ALDH1A2, coding for the retinoic acid synthesis enzyme RALDH2, identifies

Authors: C Shepherd, D Zhu, AJ Skelton, J Combe, LN Reynard, J Loughlin

Journal: Osteoarthritis and Cartilage, 2017

Expression analysis of the osteoarthritis genetic susceptibility locus mapping to an intron of the MCF2L gene and marked by the polymorphism rs11842874

Authors: C Shepherd, AJ Skelton, MD Rushton, LN Reynard, J Loughlin

Journal: Osteoarthritis and Cartilage, 2016

Methylation quantitative trait locus analysis of osteoarthritis links epigenetics with genetic risk

Authors: MD Rushton, LN Reynard, DA Young, C Shepherd, G Aubourg, F Gee, et al.

Journal: Human Molecular Genetics, 2015

Improved catenated structures of bovine peroxiredoxin III F190L reveal details of ring-ring interactions and a novel conformational state

Authors: Z Cao, DP McGow, C Shepherd, JG Lindsay

Journal: PLoS One, 2015

Methylation of cartilage DNA is a mediator of genetic risk at several OA susceptibility loci

Authors: MD Rushton, LN Reynard, DA Young, C Shepherd, R Darlay, HJ Cordell, et al.

Journal: Osteoarthritis and Cartilage, 2015

Inactivation of mammalian Ero1α is catalysed by specific protein disulfide-isomerases

Authors: C Shepherd, OBV Oka, NJ Bulleid

Journal: Biochemical Journal, 2014

Functional Characterization of the Osteoarthritis Genetic Risk Residing at ALDH1A2 Identifies rs12915901 as a Key Target Variant. Arthritis Rheumatol 70, 1577-1587 (2018).