One Health: pathogen surveillance and antimicrobial discovery

Tick-borne pathogen surveillance in Saudi livestock

Ticks feed on camels and carry bacteria inside them. Some of those bacteria infect people. What Saudi camel ticks carry — and which of those bacteria already hold antibiotic-resistance genes — is still only partly described, particularly how it varies between regions and which culturable isolates matter clinically. That gap is the project.

You would do every step yourself:

1. Collect ticks from camels in the field — Hail, Madinah and Riyadh.
2. Extract DNA from the ticks.
3. Detect which pathogens they carry, by PCR and by 16S rRNA sequencing.
4. Grow the bacteria and test them for antibiotic resistance.
5. Sequence the important ones and publish the result.

We have already completed this path once, from the field to a published genome. You would be extending work that is proven, not starting from nothing.

Hyalomma dromedariifield work includedMicroorganisms 2026;14(7):1428

Building a living model for antimicrobial resistance

To test whether a treatment works, you need a living host. Ours is a worm, Caenorhabditis elegans — small, cheap, fast, and subject to fewer ethical constraints than vertebrate models.

This stream has two halves. The first is isolating and characterising native Saudi strains of C. elegans from local soil, building a characterised local strain set available to the lab's work. The second is using that model to study infection and to test antimicrobial compounds inside a living host, not in laboratory assays alone.

What you build in this stream stays in the lab, and those who come after you will work with it.

native strain isolationfeeding & killing assays

Screening compounds for antimicrobial activity

Resistant bacteria are outrunning our antibiotics. This stream searches for what might replace them.

You would test real compounds against real resistant bacteria — plant extracts, algae extracts and silver nanoparticles, measured by MIC, MBC and biofilm assays. Our freezer holds the hard cases, including MRSA, extensively drug-resistant Acinetobacter, and one Pseudomonas aeruginosa that resisted every drug tested.

MIC · MBC · biofilm135 isolates

Computation and AI

Not every project needs a laboratory. This stream stands on its own, and suits a student who wants to publish early, or who cannot spend long hours at the bench.

What we work with today: in-silico screening of drug compounds on GPU — molecular docking, molecular dynamics, and prediction of safety and absorption. Mathematical modelling of bacterial regulatory switches. Machine learning in our clinical studies: predictive models on patient data that can show which factors contributed to the prediction. And asking new questions of large public databases that already exist.

Open questions we would welcome someone taking on: can a model be taught to tell bacteria apart from their laboratory data? Can an isolate's resistance be predicted before the susceptibility test finishes? Can a system be built that supports the laboratory specialist in a diagnostic decision?

If you have another idea in AI or laboratory informatics, tell me and we will shape the project around it.

GDC/TCGA · GEO · NCBI SRA

Studying patients from records that already exist

A retrospective study asks a new question of patient records that hospitals have already collected. You do not wait for samples, you do not wait for a sequencing run, and approved protocols exist that a project can start from, depending on the question and the institution's requirements.

That matters for a Master's. The timeline is short, the scope is defined, and a well-run project can reach a publishable draft within the degree.

Examples of the kind of study you would learn to run. These four are active in the lab, each with its own approved protocol at a Riyadh teaching hospital. You would not simply be given one — you would learn the method on it, then design your own question the same way:

1. Helicobacter pylori — infection and treatment outcome over time.
2. Urinary tract infection in children born with kidney malformation.
3. Bloodstream infection in adults.
4. Late-onset sepsis in newborn babies.

What you would learn: how to build a clean patient cohort, machine-learning prediction that can show which factors contributed to the prediction, survival analysis, and how to report a clinical study following the STROBE guidance, so the report is clear and complete.

The analysis pipeline is already built and tested. Your year is spent on your question, not on inventing the method.

Cox · Kaplan–MeierSTROBE reportingapproved protocols