Utilising statistical principles to improve design and analysis of laboratory experiments
Supervisors: Dr C I Jones, Prof S Newbury, Dr Ben Towler
Application deadline: Monday 17 March
Competition Funded PhD Project (UK Students Only)
About the Project
We are looking for an enthusiastic and motivated PhD student to join our team at Brighton and Sussex Medical School. The candidate will work closely with researchers with extensive expertise in genetics, molecular and developmental biology, gene expression measurement techniques, data analysis, and statistics (1-4).
Gene expression measurement/comparison techniques such as quantitative PCR (qPCR) and RNA-sequencing (RNA-seq) are widely used in studies involving cancer tumour profiling, biomarker identification, drug response prediction, immune cell profiling, and stem cell regeneration. Robust experimental designs and analyses are essential for generating results that are reproducible and can be used for translational or personalised medicine, and have the potential to be developed into interventions for use clinical trials. Gene expression data from patient material can be very variable, due to variations in sample collection/preparation, and genetic variation between patients. Moreover, methodologies such as long read (Nanopore) and short read (Illumina) sequencing can have their own biases due to particular chemistries involved. Therefore, the statistical and experimental design of such experiments needs to be robust to ensure results are reproducible and reflect the true underlying cellular basis of the disease/mechanism being studied. Unfortunately, many studies do not pre-define hypotheses or consider sample size/power, and are analysed with an oversimplified focus on “statistical significance”. This leads to biased results and directly contributes to the replication crisis, where the published results of many studies are unreproducible (5).
This project will, in collaboration with the NIHR Statistics Laboratory Studies group, contribute to work currently being conducted to improve design, analysis and presentation of gene expression experiments, including analysing the effect of robust experimental designs for qPCR, RNA-seq and other high-throughput methods for specific clinical/molecular biology research questions. The project will involve performing qPCR and short/long read RNA-seq experiments in the lab using state-of-the-art equipment (Illumina NextSeq and Nanopore PromethION 2 solo), statistical modelling/simulations, and bioinformatic analyses. The student will develop novel computer simulations to model real-world and ideal experimental conditions using Stata, R, or Python. The simulations will involve creating datasets representative of real populations and then drawing samples from these to simulate performing experiments with differing designs. The research carried out will be of fundamental importance in the increasing use of genomics and personalised medicine for the prognosis and diagnosis of human diseases, including cancer.
Aim 1: To perform robustly designed and powered experiments (qPCR/RNA-seq, aligned with ongoing work in the SFN/BPT labs) and use this data to quantify the effect of guidelines for robust qPCR experimental analysis by comparing different methods. Existing and simulated datasets will be used to quantify the effect of robust guidelines vs less appropriate methods to see how results and conclusions change, and how this affects published results and their reproducibility.
Aim 2: To quantify the effect of outliers and replicates in gene expression experiments (qPCR, short/long read RNA-seq). Experimental designs involve varied numbers of technical and biological replicates and varying methods are used to deal with outliers. Using new data, existing datasets, and simulations, this aim will consider how these methods affect conclusions and reproducibility, and determine the most robust approaches for each technique.
Aim 3: To assess the effect of modelling approaches and outliers on sample size/power to produce further guidance for researchers designing qPCR and RNA-seq experiments, to ensure conclusions on fundamental biological processes that are relevant to human disease. Different modelling methods (e.g., including adjustment for additional variables, mixed effects modelling etc.) will be considered. This aim seeks to leverage sophisticated statistical techniques to increase the efficiency and power of gene expression data analysis approaches, compared to commonly used simplistic techniques.
The student will be based in the BSMS Primary Care and Public Health department alongside interdisciplinary statisticians and health researchers, and gain hands-on experience in generating data using the relevant laboratory techniques in the Newbury/Towler labs. They will join the Sussex RNA group and Sussex Cancer Research Centre and collaborate with the NIHR Laboratory Studies group. The supervisory team have extensive experience in molecular techniques, statistics, and bioinformatics and work closely with clinical academics studying the genetic basis of cancers such as myeloma and glioma. The student will develop a strong understanding of statistical approaches across different research areas, with a unique understanding of biological/patient sample preparation and laboratory techniques.
Entry requirements
This studentship is suitable for those with a background in lab science or statistics (experience is not required in both). We invite applications from students who have received or are on target to achieve a relevant undergraduate degree with minimum 2:1 classification (or equivalent). Previous laboratory or statistical experience is desirable but not essential.
How to apply
Applicants must apply through the University of Brighton application Portal (StudentView) where they can submit a CV and complete the application form. The deadline for applications is 17th March 2025. Interviews will be held on 15th and 16th April 2025.
Informal enquiries are welcome and should be submitted to Dr Chris Jones: c.i.jones@bsms.ac.uk.
Funding Notes
This is a 3-year PhD studentship funded by Brighton and Sussex Medical funded, starting on 1st October 2025. Funding will cover tuition fees for UK students (at the Home rate), a stipend at the UKRI rate and a research allowance which will cover research running costs. International applicants are welcome to apply but will be required to cover the difference between Home and International fees.
References
1. Ioannidis, JPA (2005) Why Most Published Research Findings Are False. PLoS Medicine 2:e124. https://doi.org/10.1371/journal.pmed.0020124
2. Moher D, Hopewell S, Schulz KF, Montori V, Gøtzsche PC, Devereaux PJ, Elbourne D, Egger M, Altman DG. (2010) CONSORT 2010 Explanation and Elaboration: updated guidelines for reporting parallel group randomised trials British Medical Journal 340:c869 https://doi.org/10.1136/bmj.c869
3. Jones CI, Zabolotskaya MV, King AJ, Stewart HJS, Horne GA, Chevassut TJ and Newbury SF (2012) Identification of circulating microRNAs as diagnostic biomarkers for use in multiple myeloma. British Journal of Cancer, 107 (12). pp. 1987-1996. https://doi.org/10.1038/bjc.2012.525
4. Jones CI, Pashler AL, Towler BP, Robinson SR and Newbury SF (2016) RNA-seq reveals post-transcriptional regulation of Drosophila insulin-like peptide dilp8 and the neuropeptide-like precursor Nplp2 by the exoribonuclease Pacman/XRN1. Nucleic Acids Research, 44 (1). pp. 267-280. https://doi.org/10.1093/nar/gkv1336