PhD FindDoctoral opportunities · PhDFind
Menu
中文

PhD in Trustworthy AI for Predicting Tissue Gene Expression from Blood

Project 1862 · Earlham Institute · Graduate Programme

Application workspace

Save this opportunity and prepare around its actual requirements. Confirm deadlines and eligibility at the original source.

Read the research report View the recruitment source
Country or region
United Kingdom
Application deadline
2026-11-30 (confirm the exact time with the source)
Funding
Eligible international applicants can compete for NRPDTP's four-year full-tuition award, UKRI-rate stipend and £5,000 per year in research training support. International students are capped at 30% of annual programme intake, not a project quota or individual probability. £21,805 is the 2026/27 reference, not a confirmed 2027/28 amount. The project is shortlisted for funding; this does not confirm an individual award. Eligible personal visa/IHS costs may be reimbursed up to £3,500 after registration with evidence; dependants are excluded, and upfront or excess costs remain.

The research opportunity

PhD in Trustworthy AI for Predicting Tissue Gene Expression from Blood

Project overview

This Earlham Institute project, led by Dr Rahuman S Malik Sheriff, asks whether molecular signals in blood can predict gene expression in other tissues and when such predictions should be trusted. It is a four-year, full-time PhD competition within the Norwich Research Park Biosciences Doctoral Training Programme (NRPDTP). The advert says the project was shortlisted for funding. The planned start date is 1 October 2027 and the deadline is 30 November 2026, with no time or time zone stated. Current application availability has not been confirmed.

Research question and doctoral work

Existing work shows that blood can predict expression for some genes and tissues, but results are often close to population averages, accuracy varies across contexts and methods rarely identify which individual predictions are reliable. The proposed Blood2Tissue framework conditions predictions on tissue and combines expression prediction with uncertainty assessment.

The doctoral work will benchmark existing methods on GTEx, develop and calibrate a new model, and use IBDverse and reference cell atlases to investigate why some molecular expression programmes are visible in blood while others remain tissue-autonomous. The team’s CHROMATIN cross-modal attention framework imputes missing modalities in multi-omics data. MOCHA is the team’s uncertainty framework; the advert available on 19 October 2026 described its preprint as forthcoming, so its publication status and external validation remain unconfirmed. The advert also confirms visits to Ellen McDonagh’s Open Targets team at EMBL-EBI, to learn how reliable molecular signals can be connected with genetic and disease evidence.

IBDverse datasets have different access conditions. Their presence on the portal does not establish access to every donor-level file. The project aims to produce reusable computational methods and improve understanding of cross-tissue regulation; it is not an established clinical test, personal diagnosis or completed whole-body digital twin.

Host, supervision and training

Earlham is the research host, UEA is the doctoral registration and application route, and the EMBL-EBI visit does not establish a joint degree. The UEA School, degree title or code, and formal supervisory panel remain unconfirmed. Sheriff (rahuman.sheriff@earlham.ac.uk) is an Earlham Group Leader; Irene Papatheodorou contributes expertise in cellular genomics and FAIR data. The advert confirms the Open Targets visit, but does not establish Papatheodorou or the visiting team as formal co-supervisors or specify their supervision time.

See Sheriff’s current Earlham profile.

Earlham, the NBI Graduate School and Norwich Research Park provide a data-intensive bioscience setting. HPC, virtual machines, Galaxy and data support do not specify a student’s GPU, storage or staff allocation. NRPDTP includes training-needs assessment, a personal development plan and specialist training. All students complete a three-month PIPS professional internship within the four-year programme, normally outside the thesis topic; this is separate from the scientific Open Targets visit.

Eligibility, application and dates

The minimum academic requirement is a UK-equivalent 2:1 Honours Bachelor’s degree. NRPDTP may consider a 2:2 undergraduate degree where the applicant also has a relevant qualifying Master’s degree or substantial relevant research or non-academic experience; each case is assessed individually. The advert does not require a prior Master’s degree, a particular programming language or a publication.

English requirements are IELTS 6.5 overall and 6.0 in each component; TOEFL iBT 88 (reading 20, writing 19, listening 19, speaking 21); or iBT26 4.5 overall (reading, writing and listening 3.5 each; speaking 4, including Home Edition). Scores must be within the table’s two-year validity period at course start. English can be supplied under an offer condition; English-medium education in China does not automatically waive the requirement.

Apply through UEA using the required NRPDTP Personal Statement and Research Statement proformas; omission leads to rejection. Applicants may apply to up to three projects, with a separate application for each. Materials include degree certificates and transcripts, postgraduate records if applicable, a roughly 1–2-page CV, at least two referees (including one academic referee), and identity, immigration or English documents where applicable. Provide official English translations where required; confirm with the project which specific documents require translation. Proposed supervisors, friends and relatives cannot act as referees, and statements must be written by the applicant.

The advert lists interviews for 2–4 February 2027. Current application availability has not been confirmed; check with admissions before relying on the listed deadline. The partners state that selection does not distinguish on the basis of gender, marital or civil partnership status, disability, race, nationality, ethnic origin, religion or belief, sexual orientation, age or social background. To improve accessibility and attract applicants from underrepresented groups, NRPDTP requires bespoke Personal Statement and Research Statement templates, designed to let applicants present their experience and potential through suitable examples and evidence.

Funding and career development

Eligible international applicants can compete for NRPDTP’s four-year full-tuition award, a UKRI-rate stipend and £5,000 per year in research training support. International students are capped at 30% of annual programme intake; this is not a project quota or an individual probability. £21,805 is the 2026/27 stipend reference, not a confirmed 2027/28 personal amount. The project was shortlisted for funding; this does not mean that an applicant has received an award. Eligible personal visa/IHS costs may be reimbursed up to £3,500 after enrolment with evidence. Dependants are excluded, and upfront and excess costs remain.

Points to confirm

Confirm individual funding and fee terms, the 2027 stipend, visa reimbursement scope, UEA registration details, supervisory roles, CHROMATIN/MOCHA manuscripts and code, accessible datasets and donor permissions, evaluation coverage across tissues and populations, the Open Targets visit plan and personal computing allocation. Further information is available from the NRPDTP website.

Research and application brief

Report generated: . This date does not confirm that recruitment remains open.

Information updated: . Recheck the original notice before applying.

Show report contents

PhD in Trustworthy AI for Predicting Tissue Gene Expression from Blood

Quick Decision

A strong opportunity, subject to confirming the research fit. This is a computational bioscience PhD about determining when blood contains useful information about another tissue, and when a model should express doubt. Its attraction is the combination of transcriptomic prediction, individual uncertainty and biological interpretation within Earlham’s data-intensive environment. The saved official project information available on 19 October 2026 names Rahuman Sheriff as supervisor and gives a start date of 1 October 2027 and an application deadline of 30 November 2026, with no published hour or time zone. NRPDTP offers a four-year international full-fee and maintenance competition; selection remains competitive. Before applying, request the current CHROMATIN/MOCHA manuscripts and clarify available paired datasets and supervision arrangements. Project, funding.

This timing comes from the saved 19 October 2026 project information; current application availability has not been confirmed.

Research Problem and Doctoral Work

A strong average prediction can conceal an unreliable answer for a particular person, gene or tissue. Blood-to-tissue inference therefore needs to answer two questions together: what can be predicted, and how uncertainty changes across biological contexts. Blood cell composition, disease state, missing tissue observations and cohort differences may create apparent signal that transfers poorly. These are analytical risks to investigate, rather than established defects of the proposed method.

The advertised work combines GTEx benchmarking, development and calibration of a tissue-conditioned model, and biological interpretation using IBDverse and reference cell atlases. A useful thesis would distinguish donor-level generalisation from learning a tissue’s average expression pattern. Donor-disjoint evaluation, comparison with simple baselines, and scrutiny of performance across tissues and expression programmes are reasonable methodological preparation; they are not a published project protocol. Confidence estimates also need empirical calibration against held-out outcomes. Correlation alone cannot establish that an individual prediction is reliable or clinically useful.

An external reference illustrates both progress and the remaining gap. He et al.’s MTM (2023) uses multi-task learning to share information across tissues. The original indexed Methods and Results describe GTEx evaluation and an external diabetes-related analysis. This supports cross-tissue prediction as an active research direction, while differences between gene-level, sample-level and biological-association metrics matter. It does not validate Blood2Tissue or supply calibrated personal confidence. The earlier evidence assessment covered the original abstract and selected indexed Methods/Results, not the complete article, supplementary material or raw split implementation. MTM.

The consortium’s own IBDverse portal contains gut-and-blood single-cell resources with differing availability: some data are accessible, others are marked publication-dependent or restricted. Dataset names therefore do not establish immediate access to every donor-level file. The project’s CHROMATIN framework is described publicly, but an independently identifiable publication remains unconfirmed; the advert available on 19 October 2026 describes MOCHA’s preprint as forthcoming. Their manuscripts, code readiness and fallback scope need confirmation. The long-term digital-twin motivation should remain separate from a validated whole-body model or a medical test. IBDverse.

School, Programme and Field Context

Earlham is the research host; UEA is the supported doctoral registration route. Earlham’s official guidance explicitly requires candidates to meet UEA PhD registration requirements and directs studentship applications through UEA. The host and the doctoral registration institution have distinct roles. The individual UEA School, registration title/code and formal supervisory panel are still to be confirmed. A visit to EMBL-EBI does not establish an EMBL or joint degree. Earlham’s official address places the host in Norwich, UK. Registration guidance, location.

The relevant institutional advantage is a specialist genomics and computational biology environment, with NBI Graduate School support and links across Norwich Research Park. Its research computing service lists HPC, virtual machines, Galaxy and data/workflow support. These capabilities support the proposed work, but do not specify a student’s GPU quota, storage allocation or staff assistance. Ask about those concrete allocations. Postgraduate environment, computing infrastructure.

NRPDTP combines a single main research project with a Personal Development Plan, initial training-needs assessment, cohort activities and specialist development. Its three-month professional internship is within the four-year programme and is normally unrelated directly to the thesis; it is distinct from the advertised scientific Open Targets visits. Specific modules and visit duration are not yet fixed publicly. QS places UEA 381 overall in 2027; that university-wide measure does not rank Earlham, Sheriff’s new group or this methodological niche. Training access, feedback and data governance are more direct indicators for this choice. Training, programme, QS.

Supervisor and Lab Assessment

Sheriff is currently an Earlham Group Leader. The Institute announced his September 2026 arrival; older EMBL-EBI titles describe earlier employment. His official profile lists hybrid mechanistic/AI modelling, reproducible modelling infrastructure, an Imperial honorary appointment and a Florida affiliate appointment. Those external affiliations do not imply additional registration or funded travel. The advert names Irene Papatheodorou as contributing cellular-genomics and FAIR-data expertise and confirms visits to Ellen McDonagh’s Open Targets team. Formal co-supervisor status and meeting commitments remain unspecified. Current profile.

Three genuine Sheriff-coauthored original contributions clarify the team’s computational strengths:

Poincaré and SimBio (2024)

The Poincaré and SimBio (2024) study develops extensible Python tools for dynamical systems and reaction-network models, including SBML support. The original abstract and publisher-indexed Results explain a modular architecture. This is relevant to reusable computational methods and interpretable modelling, but it is not evidence of accurate blood-to-tissue inference. The full article, supplementary material and runtime benchmarks were not independently audited. Original record.

SBMLtoOdin and Menelmacar (2025)

The SBMLtoOdin and Menelmacar (2025) study translates supported model files into executable R-based simulations and interactive browser views. The complete author abstract and indexed accepted-paper passages describe filtering 1,096 curated models to 590 supported translations. That compatibility boundary is informative: a reusable tool can still have a restricted domain. The April preprint and September Bioinformatics article are one work, not two independent results. Author preprint.

BioSimulators (2022)

The BioSimulators (2022) study introduces a registry, consistent interfaces and standards-based validation for simulation tools. The earlier evidence assessment covered the original institutional manuscript’s indexed abstract and Introduction. Its historical tool counts demonstrate implementation scale at publication, not current coverage or biological predictive validity. Sheriff is a collaborator in a large author team, not inferred to be sole lead. Institutional manuscript.

The available evidence has unequal reading depth across these three originals; it does not include three complete full texts. Together they support reproducibility and software infrastructure experience. Public evidence is thinner on the exact new uncertainty framework. A newly established group also warrants questions about day-to-day mentoring, continuity, student numbers and manuscript/code availability.

Funding, Application, and Living Cost

The recruitment description identifies the project as shortlisted for funding, which is distinct from an individual award. NRPDTP policy available in October 2026 explicitly accepts international candidates for fully funded studentships, covering four-year tuition, a UKRI-rate stipend and £5,000 annual research training support. The research grant is not personal living income. International appointments are capped at 30% of annual programme intake, not 30% admission probability or a project-specific quota. £21,805 is the published 2026/27 stipend reference, not a confirmed personal 2027/28 amount. For October 2027 entrants, eligible individual visa/IHS expenditure can be reimbursed up to £3,500, after enrolment and registration with evidence; dependants are excluded. Initial cash flow can therefore still matter. Funding, eligibility.

Applications use the UEA system and the mandatory NRPDTP Personal Statement and Research Statement proformas. Omission causes rejection. The guidance permits up to three projects, each with a separate application, and asks for degree certificates and transcripts, postgraduate records if applicable, a roughly one-to-two-page CV, at least two referees and relevant identity, immigration and English documents. Provide official English translations where required; confirm with the project which specific documents require translation. At least one referee must be academic; intended supervisors, friends and relatives are excluded. This bespoke route takes precedence over a generic research-proposal template. The timetable confirms 30 November 2026, interviews on 2–4 February 2027, and 1 October 2027 commencement. The deadline hour and time zone remain unconfirmed. Application, dates.

The partners state that selection does not distinguish on the basis of gender, marital or civil partnership status, disability, race, nationality, ethnic origin, religion or belief, sexual orientation, age or social background. To improve accessibility and attract applicants from underrepresented groups, NRPDTP requires bespoke Personal Statement and Research Statement templates so applicants can present their experience and potential through suitable examples and evidence.

UEA’s 2026/27 accommodation schedule gives Village rooms at £178.50 weekly and Crome rooms at £209.50, with 41-week totals of £7,318.50 and £8,589.50. Extending those weekly rates over 52 weeks gives an illustrative £9,282–£10,894 rent scenario; availability and 2027 prices are unconfirmed. Against the dated £21,805 stipend reference, this leaves roughly £10,911–£12,523 before food, transport and other costs. It is not a promised disposable-income calculation. Housing prices.

October 2026 government references list a £558 Student visa application fee and £776 student IHS per visa year; actual visa length and charging rules govern the total. The outside-London maintenance evidence figure is £1,171 monthly for up to nine months. China’s differential-evidence status can simplify initial documentation, but does not remove the financial requirement or prevent later evidence requests. Visa, financial evidence, IHS.

Applicant Fit and Preparation

The project specifies a UK 2:1 honours degree or equivalent and IELTS 6.5 overall, each component at least 6.0. It does not publish a mandatory Master’s degree, particular programming language or prior paper requirement. Relevant preparation could come from quantitative bioscience, statistics, machine learning or bioinformatics, provided the applicant can engage with transcriptomic biology. A purely software-focused application without biological reasoning, or an expectation of immediate clinical deployment, would miss important parts of the work.

There is a positive Chinese-applicant route: NRPDTP accepts equivalent overseas qualifications and international funding applications; UEA’s Chinese guidance states that holders of a recognised university Master’s degree may be considered for doctoral study. Its Chinese undergraduate percentages for taught Master’s entry must not be reused as a PhD cutoff. The DTP’s lower-second exception requires a relevant qualifying Master’s or substantial relevant experience; an individual qualification assessment is still necessary. Eligibility, Chinese guidance, general PGR requirements.

The Science/Medicine research English table available in October 2026 also lists TOEFL iBT 88, with reading 20, writing 19, listening 19 and speaking 21; its iBT26 scale lists 4.5 overall, with 3.5/3.5/3.5/4 respectively. These test results and IELTS must remain within their two-year validity window at course start. The programme allows language evidence to become an offer condition. English-medium study in China does not automatically establish an exemption. Other accepted qualifications and any waiver should be checked against the offer. English equivalencies.

Suggested preparation is a small, reproducible computational analysis showing an honest train/test separation, sensible baselines and interpretation of uncertainty. This is an analyst’s suggestion, not an additional application requirement. A concise statement should explain the biological question, relevant evidence of quantitative work and specific training needs. Applicant-produced statements are required; AI-assisted research coding in the advert does not authorise outsourced application writing.

Next Actions and Questions to Verify

Before Applying

Contact rahuman.sheriff@earlham.ac.uk for the current CHROMATIN and MOCHA manuscripts, code status, core thesis question and intended validation datasets. Confirm the UEA School/registration selection, exact application time zone, mandatory template versions and reference arrangements. The official application page links some filenames bearing older years; confirm that the linked versions apply to the 2027 round.

At Interview

Ask who provides daily methodological and biological supervision; which paired donor datasets and restricted resources are already approved; how calibration will be tested under tissue and disease shifts; and what a defensible thesis would contain if some tissues are poorly predictable. Confirm HPC/storage support and how visits to Open Targets fit the research timetable. Neither the project’s longer-term motivation nor its collaborator names establish clinical utility or a placement job.

Before Accepting

Obtain written four-year fee/stipend terms, the 2027 rate, payment dates, visa reimbursement process, registration details and any continuation costs. The absence of an individual award before competition is not itself a reason to exclude this opportunity: the international full-funding framework is explicit. The main unresolved issues concern the new frameworks’ availability, data access and supervision in a new group. Recruitment information reflects 19 October 2026 sources; later availability should be confirmed with admissions.

Career development is credible in computational biology and research software because the programme includes a professional internship and the local Institute advertises roles demanding reproducible software, evaluation and data/compute integration. That is evidence of relevant skill categories, not a forecast of salary or employment. The scientific Open Targets visit and the unrelated professional PIPS should be planned as separate training activities. PIPS, research software role.

Evidence and Sources

Project, institution and supervision

Original research and data

Funding, application and living conditions

Sources

The listing may come from a university or recruitment platform. Check the notice and official sources in the report for current application requirements. View the recruitment source

Explore opportunities with related research or the same country, prioritising similar funding where relevant. Recheck each notice before applying.

University of Sheffield · United Kingdom

Bio-Inspired MEMS PhD: Acoustic Sensing for Buried Water-Pipe Condition Assessment

Competitive EPSRC Doctoral Landscape Award. The project listing advertises 3.5 years, full tuition, a UKRI-rate stipend currently stated as £21,805/year (2026/27 benchmark), and a project research grant. EPSRC funding is based on the Home tuition rate; Sheffield says it covers the difference for successful international awardees. International awards are capped at 30%, with up to ten places anticipated by Sheffield and Home applications considered first. The 2027/28 stipend and project-grant amount/scope are not stated.

2027-01-13 (confirm the exact time with the source)

University of Leeds · United Kingdom

Landscape Ecology PhD: Connectivity and Ecosystem Function Along Canals

YES-DTN 2027/28 competitive studentship. Scheme-level terms in project materials: 3.5 years, UKRI-rate stipend, university fees, and research/fieldwork/training/conference-travel support. International applicants may compete; up to 30% of awards may go to international candidates, with full applicable Home or International university tuition for successful award holders. Relocation, visa and NHS surcharge costs are excluded. Project-specific stipend amount and fieldwork allocation are not stated.

2027-01-12 (confirm the exact time with the source)

University of Leeds · United Kingdom

Satellite Remote Sensing PhD: Comparing Soil Organic Carbon and Moisture under Regenerative and Conventional Farming

A 3.5-year YES-DTN award is advertised with a UKRI-rate stipend, university fees, and research, fieldwork, training and conference support. The 2027/28 stipend, project research-cost allowance, 2027 international allocation and project-specific international eligibility are unknown. Visa, immigration health surcharge and relocation costs are excluded.

2027-01-12 (confirm the exact time with the source)

University of Edinburgh · United Kingdom

PhD in evolutionary cell biology | How does Diplonema papillatum segregate its chromosomes?

Listed as a competition-funded PhD opportunity for applicants worldwide. The Darwin Trust route is for non-UK nationals whose first university degree or equivalent was obtained less than five years before the closing date; the studentship lasts up to four years and includes an annual stipend and University tuition fees. The stipend amount and full overseas-fee coverage remain unconfirmed. A 2026–2027 overseas fee reference is £34,800 per year; the 2027–2028 fee and any shortfall are unconfirmed.

2027-01-06 (confirm the exact time with the source)