How-to articleResearch design

Runs-of-Homozygosity Analysis for Recessive Disease Burden and Expanded Carrier Screening Panels in the UAE: A PhD Research Design Blueprint

Overview

The UAE already screens. Mandatory premarital genetic testing for Emiratis began in early 2025, with MoHAP describing a 570-gene panel covering more than 840 conditions [1]. Subsequent Dubai implementation research has since evaluated a 782-gene sequencing panel [15], so the operational and evidentiary landscape is already evolving. That changes the research question. A PhD proposal can no longer treat the basic implementation of premarital genetic screening in the UAE as an unresolved policy question, and the contribution of consanguinity to recessive risk is well established in the published literature. The more specific research opportunity is narrower and more technical: which genes belong on the panel, at what carrier-frequency threshold, and what genomic evidence should decide that? This blueprint shows how the research question can be translated into a defensible doctoral research design.

Who this is for: PhD applicants and scholars in medical genetics, genetic counselling, bioinformatics, public-health genomics and paediatrics; clinician-researchers; and supervisors shaping candidate projects in the UAE and wider GCC.

Scope and separation of content. This article distinguishes three things throughout: published evidence, cited to source; research hypotheses, which a thesis would test and which are labelled as such; and design guidance from PhD Assistance. It is research-methodology guidance only — not clinical, diagnostic or genetic-counselling advice, and not guidance for anyone’s reproductive, marital or family decisions. It does not assess any individual’s or couple’s risk. Decisions about national screening programmes rest with health authorities. Regulatory and data-access requirements change; confirm them with your research ethics committee and the relevant health regulator before submission.

A note on framing. Consanguineous marriage is a longstanding social practice in the region, and research on it carries real potential for stigma. Throughout this blueprint, consanguinity is treated as a population-genetic variable to be measured, not a behaviour to be evaluated. Section 10 sets out why that framing is also an ethics-committee requirement rather than only a courtesy.

1. Why recessive disease burden is a high-value PhD research area in the UAE

The policy infrastructure now exists, so the research question must now address evidence gaps within that system. Mandatory premarital genetic testing for Emiratis began in January 2025, covering 570 genes and over 840 conditions [1]. Federal Decree-Law No. 49 of 2023 makes premarital genomic screening obligatory, requires evidence of screening before a marriage contract is completed, and provides that the list of genetic diseases to be examined is issued by ministerial resolution (Art. 15) [2]. Because the list is established by ministerial resolution, its composition can be changed through the relevant regulatory process; population-specific genomic evidence is one source of evidence that can inform such an evaluation.

The genomic substrate is documented. A 2025 medRxiv preprint, not yet peer-reviewed, analysing 43,608 Emirati Genome Programme genomes reported elevated median inbreeding coefficients and cumulative runs-of-homozygosity length, largest in groups with Arabian main ancestry and higher than values reported for Qatar, and reconstructed 247 families with third- or fourth-degree consanguineous parents [3]. UAE consanguinity has been reported in the 39–54% range across studies [4].

The disease catalogue is population-specific. The Catalogue for Transmission Genetics in Arabs records 665 distinct genetic conditions described in Emiratis, 621 following clear Mendelian inheritance, with 1,365 gene variants reported. Over half are extremely rare globally and predominantly autosomal recessive. The same analysis found 235 variants absent from dbSNP and ClinVar, and 307 CTGA-pathogenic variants classified differently in ClinVar [5]. Population-specific interpretation is therefore not a theoretical concern.

A regional template exists — and it is not Emirati. In 6,045 Qatari whole genomes, 62.5% of participants carried at least one known pathogenic variant for a recessive condition; homozygosity for a recessive pathogenic variant occurred in roughly 1 in 150 subjects; Peninsular Arabs showed 5.8-fold enrichment for homozygous pathogenic variants relative to other ancestries, and 56 genes exceeded a carrier frequency of 1/50 [6]. That study demonstrates the design is achievable in a Gulf biobank. It also shows what an Emirati equivalent would need to match.

What changed in the UAE evidence base during 2025–2026. Two studies published since the panel launched reset the baseline. Jain et al. analysed 1,194 Emirati exomes alongside 1,333 Arab Emirati families, reported carrier rates including CYP21A2 at 10.6%, HBB at 9.6%, MEFV at 5.9% and ABCA4 at 4.3%, estimated at-risk couple rates of 4–21% depending on the panel assumed, and proposed a provisional Emirati premarital screening gene list of 481 genes — noting that 214 of their 425 additional genes (50%) are not on the 570-gene MoHAP list [14]. Alblooshi et al. then reported a citywide Dubai programme that screened 1,000 couples on a 782-gene panel, of which 558 genes were MoHAP-mandated and 224 were added from a local Emirati clinical genomic database; 79 couples (8%, 95% CI 6.4–9.7) were identified as at risk, and 14 of those 79 (18%) were detected only through the locally sourced genes. Consanguinity or shared ancestry met their threshold in 28.0% (160/572) of whole-genome-sequenced couples [15]. Emirati carrier-frequency estimation is therefore no longer the open question.

The research gap, stated narrowly. Runs of homozygosity analysis in this setting is not an unexplored idea; autozygosity has been characterised in Emiratis at population scale [3], and carrier-burden methods have been demonstrated regionally [6]. A defensible 2026 gap is tighter than “study consanguinity in the UAE”. It is closer to this: the remaining opportunity is not simply to estimate Emirati carrier frequencies, because recent Emirati studies have already produced gene-level carrier estimates and evaluated premarital panel composition. A narrower gap is whether population-scale ROH and FROH patterns, ROH length classes and genomic substructure can be integrated with population-specific carrier-frequency estimates into a prespecified, reproducible framework for evaluating panel composition and threshold sensitivity. Confirm any such statement with a documented systematic search — databases, search strings and date recorded — before submission.

2. Consanguinity is not the same as autozygosity: what genomes measure

This distinction is the conceptual foundation of the whole project, and proposals that blur it tend to be sent back.

Consanguinity is a reported relationship between parents — first cousins, second cousins, or a shared tribal lineage. It is collected by questionnaire, and it yields an expected inbreeding coefficient (F) from pedigree assumptions.

Autozygosity is the observed genomic consequence: the proportion of an individual’s genome present in two identical copies inherited from a common ancestor. It is measured, not reported.

The two diverge for reasons that matter methodologically:

  • Pedigree F is an expectation, not a realisation. Because recombination is stochastic, two full first cousins can have genuinely different realised autozygosity. Pedigree-based F assigns them the same value; the genome does not.
  • Reported relationships are incomplete. Endogamy over many generations produces background relatedness that no one reports as “consanguineous marriage”, because the common ancestor is too distant to be named.
  • Reporting is socially shaped. In a context where the practice is discussed publicly, self-report may be accurate; in others it may not be. Genomic measurement does not depend on disclosure.

The research implication. A thesis that uses self-reported consanguinity as its exposure variable is measuring something different from one that uses FROH. Say which you are using, justify it, and — where both are available — report the correlation between them. That comparison is itself a publishable finding in a population where endogamy is longstanding.

Hypothesis to label as such: that FROH captures recessive-risk-relevant autozygosity in Emiratis better than self-reported relationships do. It is testable in any dataset holding both, and it should be framed as a hypothesis, not an assumption.

3. What runs of homozygosity are and how they are called

A run of homozygosity is a contiguous stretch of the genome where an individual carries identical haplotypes on both chromosomes. Long ROH arise when the two copies descend from a recent common ancestor, because recombination has had few generations to break them up. The general framework, including length-class interpretation, is set out in the standard review of the field [7].

The ROH analysis pipeline. The figure below sets out the seven stages and, critically, the decision each one forces. Name this pipeline as the study’s conceptual framework in the proposal: every later claim depends on these choices being fixed in advance and reported in full.

Recreation Image 1

The ROH analysis pipeline: seven stages from genotype data through quality control, ROH calling, autozygosity metrics, homozygosity mapping and carrier frequency estimation to panel evidence, with the decision to prespecify and the output at each stage

Calling approaches to name and justify:

ApproachHow it worksStrengthsLimitations to state
Sliding-window (PLINK –homozyg) [11]Scans fixed-size SNP windows and calls ROH where a threshold proportion of windows are homozygousWidely used, well documented, results comparable with much of the literatureSensitive to window and density settings; performance depends on marker density
Model-based hidden Markov (bcftools/RoH) [12]Models homozygous versus heterozygous states along the genome, using allele frequenciesHandles genotype error probabilistically; suits sequencing dataNeeds an appropriate allele-frequency reference, which is itself a population-specific choice
Likelihood-based classification (GARLIC) [13]Classifies ROH into length classes using a likelihood framework with LD-aware windowsExplicit length-class assignment; designed for cross-population comparabilityLess commonly used, so fewer directly comparable published results

Array versus sequencing. Array data give even marker spacing and are well suited to sliding-window calling, but density limits the detection of short ROH. Sequencing data give far denser coverage but require careful handling of variable depth and genotype quality, since apparent homozygosity can arise from low coverage rather than from shared ancestry. State which you have, and what that platform can and cannot resolve.

4. Parameter choices that change ROH results — and how to justify them

This section is where a methods chapter earns its credibility. ROH results are not parameter-free observations; they are the output of a specified procedure.

ParameterWhat it controlsEffect if set too permissivelyEffect if set too stringently
Minimum ROH lengthThe shortest segment countedShort LD-driven homozygous tracts inflate FROHGenuine recent autozygosity is missed
Window size (SNPs or kb)The scanning unitFragmented calls; boundary imprecisionSmoothing over real segment breaks
Minimum SNP count and densityHow much evidence a call requiresCalls in sparsely covered regionsSystematic loss in low-density regions
Heterozygote allowance per windowTolerance for genotyping errorMerges distinct segments across true heterozygosityGenuine long ROH split by single errors
Missing-call allowanceTolerance for missingnessCalls driven by missing dataLoss of segments in imperfect samples
LD pruningWhether correlated markers are thinnedRetained LD inflates short-ROH countsPruning can remove the signal long-ROH detection depends on

How to justify choices, in four steps a supervisor will accept:

  1. Anchor to precedent. State which published parameter set you are following and in which population it was used, so your results are comparable to something.
  2. Match parameters to platform. Density thresholds appropriate for a dense sequencing dataset are not appropriate for a sparser array, and vice versa.
  3. Run a sensitivity analysis. Report FROH under at least two or three plausible parameter sets, and show whether your substantive conclusions change. In our experience reviewing genomics proposals, this single addition converts a contestable methods chapter into a defensible one.

Prespecify. Fix the primary parameter set in the protocol before analysis; report the others as sensitivity analyses rather than selecting the set that produces the most striking result.

5. FROH and ROH length classes: reading recent and ancient inbreeding

FROH is the ratio of total ROH length to the length of the autosomal genome considered. Its apparent simplicity hides a reporting trap: the denominator is a choice. Studies differ in whether they use the full autosomal length, the mappable or callable genome, or the span actually covered by the genotyping platform. Two studies reporting “FROH” may therefore not be comparable. State your denominator explicitly, and when comparing with published values, check theirs.

Length classes carry the temporal signal [7]. Short segments reflect ancient, population-level background relatedness accumulated over many generations, sustained by longstanding endogamy. Long segments reflect recent parental relatedness, because recombination has had few generations to break them. Medium segments sit between the two.

Why this matters for panel design — and it is the analytic heart of the project. The two signals imply different screening consequences. Background autozygosity from ancient endogamy tends to raise homozygosity risk broadly across a community, in ways a population-level panel can address. Recent parental relatedness concentrates risk within specific families and may raise rare, family-specific variants that no fixed panel will capture. A thesis that reports only a single pooled FROH cannot separate these, and therefore cannot say which screening strategy its own data support.

Report, at minimum: the FROH distribution rather than the mean alone; FROH by length class; segment counts alongside total length; and the denominator and parameter set used to derive all of them.

6. Homozygosity mapping: from shared ROH to candidate recessive variants

Homozygosity mapping exploits a simple logic: in a recessive condition arising from a shared ancestral variant, affected individuals will be homozygous across the surrounding region, and that region will be shared among them.

Design requirements to state in the protocol:

  1. The affected sample. How many affected individuals, related or unrelated, and with what phenotype definition. A precise, documented phenotype is as important here as the genomics.
  2. The sharing criterion. How much overlap between individuals’ ROH constitutes a shared region, and whether unaffected relatives are used to exclude regions.
  3. Prioritisation within candidate regions. How variants are ranked — by predicted consequence, gene–phenotype plausibility, frequency in population databases, and segregation where family data exist.
  4. The confirmation step. What would count as sufficient evidence: segregation in additional family members, independent cohort replication, or functional data. Say explicitly which is achievable within the project and which is not.

A limitation to state plainly. Homozygosity mapping localises regions; it does not by itself establish causation. In a highly autozygous population, individuals carry many long ROH, so candidate regions will be numerous and many will be incidental. Multiple-testing and prioritisation strategy therefore need to be prespecified, and the expected number of candidate regions should be anticipated rather than discovered.

7. Estimating carrier frequencies and classifying pathogenic variants

Carrier frequency is one of the key quantities relevant to panel evaluation, and estimating it well is mostly a classification problem.

The classification step governs everything downstream. Variants should be classified using an established clinical framework such as the ACMG/AMP criteria [8]. The decisive design question is what you count as a carrier: pathogenic variants only, or pathogenic plus likely pathogenic, and how variants of uncertain significance are handled. Each choice produces a different carrier frequency from the same data. Prespecify the rule, and report frequencies under the alternative rule as a sensitivity analysis.

Population-specific variant interpretation is therefore an important methodological issue. With 235 Emirati variants absent from major international databases and 307 classified differently between CTGA and ClinVar [5], classification decisions taken against international references alone may systematically misestimate Emirati carrier frequencies. Document your reference sources and their versions with access dates.

Statistical reporting. Report gene-level carrier frequencies with confidence intervals, prespecifying the interval method — Wilson or Clopper–Pearson rather than the normal approximation, which performs poorly for the rare frequencies that matter most here [9]. Size the study for the precision needed at your lowest decision-relevant frequency, not at the commonest.

Founder variants deserve separate treatment. The Qatari analysis identified 59 potential founder alleles with elevated frequency in subpopulations [6]. Founder variants can reach locally high frequencies while remaining invisible in pooled national estimates, which makes them relevant to evaluating panel composition — and to Section 9’s substructure analysis.

8. Which genes belong on an expanded premarital panel? Building selection criteria

This is where the thesis can move from population description toward evidence relevant to panel evaluation.

The existing international benchmark, and why it does not settle the question. The ACMG practice resource defines carrier screening tiers by carrier frequency, recommending Tier 3 — conditions with a carrier frequency of 1/200 or greater — for all pregnant patients and those planning pregnancy, with Tier 4 (below 1/200) considered for consanguineous relationships or where personal or family history warrants it [10]. Two features make this a research opening rather than an answer. First, the thresholds were derived largely from populations in which consanguinity is uncommon. Second, ACMG itself flags consanguinity as the situation in which the standard tier is insufficient — without specifying what should replace it. The ACMG resource is a United States practice framework, so its components do not all transfer equally: the tier logic and the selection criteria it applies are reusable, whereas the 1/200 cut-off and the gene lists behind it would need validation against Emirati carrier frequencies [14] before being treated as locally applicable.

The design question, stated for a proposal: what carrier-frequency threshold, applied to Emirati population genomic data, produces an expanded carrier screening panel that is defensible on detection, cost, counselling burden and equity — and how does that threshold compare with 1/200?

Five criteria a defensible gene-selection framework must operationalise:

CriterionWhat it meansHow to operationalise itWhere judgement enters
Carrier frequencyHow common the carrier state is in the target populationGene-level estimates with confidence intervals from local dataChoosing the threshold, and whether it is national or substructure-specific
SeverityThe clinical impact of the conditionA published severity classification applied by more than one rater, with agreement reportedSeverity scales embed value judgements; state which scale and why
Penetrance and expressivityWhether and how consistently carriers of two variants are affectedEvidence from the literature and local clinical cohortsPenetrance is often poorly characterised for rare conditions
ActionabilityWhether identification changes anything — reproductive options, surveillance, early treatmentExplicit categories, prespecifiedWhat counts as actionable is context- and values-dependent
Test performanceDetection rate, residual risk, and whether variants are detectable by the assayAnalytical validity for the specific variant classes in the populationAssay limits differ for structural, repeat and deep-intronic variants

Counselling burden as a sixth dimension. Expanding the number of genes screened increases the volume of carrier findings and may also increase the number of results requiring additional interpretation, depending on variant spectrum, assay design and classification uncertainty. A panel optimised on detection alone can exceed what a counselling service can deliver, which is a real constraint on implementation and belongs in the analysis — not in the discussion as an afterthought.

9. Regional and tribal substructure: why national averages can mislead

A single national carrier frequency is an average over a structured population, and averaging is exactly what can hide the cases a panel exists to catch.

The evidence that substructure matters here. Emirati autozygosity is not uniform: ROH burden was largest in groups with Arabian main ancestry within the national cohort [3]. Regionally, Peninsular Arabs showed 5.8-fold enrichment for homozygous pathogenic variants relative to other ancestry groups in the Qatari analysis [6]. For example, a hypothetical variant observed at 1/300 nationally could have a substantially higher frequency in one community and be absent or very rare elsewhere.

How to handle it analytically:

  1. Define strata genomically, not only administratively. Use genetic principal components or clustering to define ancestry groups, and report how these relate to reported region or lineage — rather than assuming the two coincide.
  2. Report stratum-specific frequencies with intervals. Expect wide intervals in small strata and say so; precision, not point estimates, determines what can be concluded.
  3. Prespecify the decision rule. State in advance whether a gene qualifies on national frequency, on any stratum exceeding the threshold, or on a weighted rule — and justify the choice.
  4. Handle stratum identification carefully. Reporting that a named tribe or region carries elevated risk has consequences beyond the thesis. Section 10 covers the governance of this directly; the analytic point is that the reporting granularity should be decided with the ethics committee before analysis, not after.

The genuine trade-off to discuss. A single national panel is simpler, cheaper and equitable in access, but may under-serve high-burden subgroups. A stratified panel may detect more but raises questions of how strata are assigned, whether assignment is acceptable to communities, and what happens to people of mixed or unassigned background. A thesis that presents the evidence for both, rather than advocating one, is stronger.

11. PhD Assistance Carrier-Panel Gene Prioritisation Framework

The framework assesses candidate genes against the five selection criteria plus counselling burden. Its primary output is a multi-criteria evidence matrix, not a single composite score: any weighting applied across criteria is itself a value judgement that would need separate justification. A composite score may be reported as a secondary, clearly labelled analysis.

DimensionEvidence sourceSuggested scoring approachPrespecify in the protocol
Carrier frequency in target populationLocal population genomic data; stratum-specific estimatesBanded against the chosen threshold, with the confidence interval carried throughThe threshold; whether national or stratum-specific qualification applies
SeverityPublished severity classification, applied by independent ratersOrdered categories, with inter-rater agreement reportedWhich classification scale; how disagreements are resolved
Penetrance and expressivityLiterature and local clinical cohortsHigh / moderate / uncertain, with the evidence basis recordedHow “uncertain” genes are treated at the decision stage
ActionabilityPrespecified categories covering reproductive options, surveillance and treatmentOrdered categoriesThe category definitions, fixed before scoring
Test performanceAssay validation data for the relevant variant classesDetectable / partially detectable / not reliably detectableHow residual risk is reported per gene
Counselling burdenExpected carrier and uncertain-result volume from the frequency dataModelled results per 1,000 screenedWhether burden acts as a weight or as a reported consequence

Worked operationalisation. The dimensions become usable only once the bands are written down before scoring any gene. One workable set, offered as a starting point to adapt rather than a validated instrument: carrier frequency scored 0–4 against the chosen threshold, from 0 below half the threshold to 4 at four times it; severity 0–3 on the chosen published classification; penetrance 0–3, with genes whose penetrance is uncertain scored 0 rather than dropped; actionability 0–3 on the prespecified categories; and test performance 0–3 on detectability for the relevant variant classes. Counselling burden is reported alongside the matrix rather than scored into it, because it is a consequence of the resulting list rather than a property of any single gene. Four rules need fixing in the protocol in advance: how missing evidence is handled — a prespecified default score with the gap recorded, rather than exclusion, which silently favours well-studied genes; how ties are broken, for example by carrier frequency with its confidence interval and then by severity; what inclusion threshold applies if a composite score is used at all; and how the result moves under at least two alternative weightings.

Keep the matrix primary, the score secondary. The defensible primary output is the full scored matrix, with every dimension visible per gene and every missing or uncertain entry flagged. A composite score is a secondary summary — useful for communication and for sensitivity analysis — and it should not be the only thing reported, because collapsing five dimensions into one number conceals exactly the judgement calls an examiner will want to inspect.

How to use it. Score every candidate gene on all six dimensions, report the full scored matrix as a thesis appendix, then present any ranked list under your prespecified threshold alongside at least one alternative threshold. Presenting the ranking’s sensitivity to the threshold is more defensible than presenting a single list, and it is more useful to anyone making the actual decision.

What the framework is not. It is a research instrument for generating and structuring evidence. It does not constitute a recommended panel, and health authorities that set the list by resolution [2] decide any changes to a national screening programme.

12. How to turn this blueprint into an approvable PhD proposal

A PhD research proposal on this topic must state the research questions, one primary aim with a measurable endpoint, the evidence gap, the expected limitations and the contribution, and pair every data dependency with a documented fallback.

Evidence gap map

Existing evidencePopulation and dataMain contributionRemaining gap
Mousa et al. 2025 [3]43,608 Emirati genomesPopulation-scale autozygosity: inbreeding coefficients, ROH length, consanguineous family reconstructionAutozygosity characterised, but not linked to gene-level carrier frequencies or panel evidence
Bizzari et al. 2023 [5]CTGA records for Emiratis665 conditions, 1,365 variants; 235 absent from international databasesA catalogue of reported variants, not population carrier frequencies
Jain et al. 2025 [14]1,194 Emirati exomes; 1,333 Arab Emirati familiesGene-level carrier rates (CYP21A2 10.6%, HBB 9.6%, MEFV 5.9%, ABCA4 4.3%); at-risk couple rates of 4–21% depending on the panel assumed; a provisional 481-gene list, of which 214 of 425 additional genes (50%) are absent from the 570-gene MoHAP panelCarrier frequencies published without linkage to autozygosity, substructure stratification or threshold sensitivity analysis
Alblooshi et al. 2026 [15]1,000 Dubai couples screened on 782 genes79 at-risk couples (8%, 95% CI 6.4–9.7); 18% of at-risk couples detected only through locally sourced genes; consanguinity or shared ancestry in 28.0% of sequenced couplesImplementation outcomes for one city panel; the gene-selection evidence and the threshold logic behind panel composition remain separate questions
Aamer et al. 2024 [6]6,045 Qatari whole genomes62.5% carriers; 1 in 150 homozygous; 56 genes above 1/50 carrier frequency; 59 founder allelesQatari, not Emirati; demonstrates the method is achievable regionally
MoHAP 2025 policy [1]; Decree-Law 49/2023 Art. 15 [2]National programmeA 570-gene mandatory panel, revisable by ministerial resolutionThe published evidential basis for gene inclusion and thresholds is not established in the public literature
ACMG practice resource [10]InternationalTiered thresholds; Tier 3 at 1/200Thresholds derived largely from non-consanguineous populations; Tier 4 flagged but unspecified for consanguinity

Novelty matrix

ComponentWhat is already knownWhat remains unknownWhat this PhD would add
AutozygosityElevated ROH and inbreeding coefficients in Emiratis, varying by ancestry group [3]FROH by length class linked to carrier burden; relationship between reported consanguinity and measured autozygosityA length-class-resolved autozygosity profile tied to recessive-risk outcomes
Carrier burdenEmirati gene-level carrier rates and a provisional 481-gene list [14]; regional estimates from Qatar [6]; an Emirati variant catalogue [5]Whether those estimates hold under interval estimation, a stated classification rule and stratification by genomic substructureCarrier frequencies regenerated under a documented classification rule, with confidence intervals and stratum-level estimates
Panel evidenceA 570-gene national panel [1]; a 782-gene Dubai implementation panel evaluated in 1,000 couples [15]; international tiers [10]Which threshold is defensible for a population with this autozygosity profile, and how sensitive panel composition is to itA criteria-scored evidence matrix with threshold sensitivity analysis, any ranking reported as a secondary view
SubstructureAncestry-group differences reported [3, 6]; consanguinity or shared ancestry met the programme threshold in 28.0% of sequenced Dubai couples [15]Whether national frequencies mask stratum-level qualifying genesStratum-specific frequencies and a prespecified qualification rule

Research questions to adapt:

  1. RQ1: What is the distribution of FROH in the study population, by ROH length class, under a prespecified calling parameter set?
  2. RQ2: How does measured autozygosity relate to reported parental consanguinity, where both are available?
  3. RQ3 (primary): What are the gene-level carrier frequencies for recessive conditions in this population, with confidence intervals, under a prespecified variant-classification rule?
  4. RQ4: Applying prespecified selection criteria, which genes qualify for an expanded panel, and how sensitive is that list to the carrier-frequency threshold?
  5. RQ5 (exploratory): Do genomic strata show carrier frequencies that differ materially from the national estimate, and would any gene qualify on stratum-specific but not national frequency?



Expected limitations to state in the proposal:

  • Sampling frame limits generalisation; a clinical or screening-derived sample is not a population sample.
  • ROH calls depend on platform and parameters; sensitivity analyses mitigate but do not remove this.
  • Carrier frequencies depend on the classification rule, and variants of uncertain significance are a substantial category in an understudied population [5].
  • Severity and actionability scoring embeds value judgements, however transparently applied.
  • Homozygosity mapping localises regions but does not establish causation without confirmation.
  • Small genomic strata yield imprecise estimates.
  • Access, consent scope and publication approvals may restrict which outputs can be released [2].
  • Panel evidence is not a panel decision, and no implementation outcome can be evaluated within a doctoral timeline.

Contribution statement. A finished thesis based on this design will provide (1) an autozygosity profile by length class for the study population with a reproducible ROH analysis pipeline; (2) gene-level carrier frequency estimates with confidence intervals for a pre-specified classification rule; (3) a criteria-scored evidence matrix covering every candidate gene, with any ranked list reported as a secondary, threshold-dependent view rather than as a panel recommendation; and (4) a prioritisation framework which can be exported to other Gulf population genetics research.

Design weaknesses that get genetics proposals sent back: novelty asserted rather than demonstrated against current policy and literature; consanguinity and autozygosity used interchangeably; ROH parameters unstated or unjustified; no sensitivity analysis; carrier frequencies reported without a classification rule or confidence intervals; national averages presented with no substructure analysis; reporting granularity for tribal or regional findings not agreed with the ethics committee in advance; no return-of-results pathway; and panel recommendations framed as policy rather than as evidence.

Frequently asked questions

How can an ROH-based PhD remain novel when Emirati carrier-frequency studies already exist?

Emirati gene-level carrier frequencies have already been published [14], and a 782-gene Dubai panel has been evaluated in 1,000 couples [15]. Novelty therefore has to come from integration: linking ROH and FROH patterns, length classes and genomic substructure to carrier burden within a prespecified, reproducible panel-evaluation framework. Confirm that this combination remains unpublished through a documented systematic search.

Is ROH analysis alone enough for a PhD?

Rarely. Autozygosity has already been characterised at population scale in Emiratis [3]. ROH analysis works as the foundation, but the doctoral contribution usually comes from linking it to gene-level carrier frequencies and to a transparent, criteria-based panel evidence framework.

Which of the three areas makes the best primary aim?

Which parameters should I use for ROH calling?

Should I use FROH or reported consanguinity as my exposure?

They measure different things, so the answer depends on the question. FROH measures realised autozygosity; reported relationship measures a social variable with an expected inbreeding coefficient attached. Where both are available, reporting their relationship is itself a finding worth publishing in an endogamous population.

How do I decide the carrier-frequency threshold?

Treat it as a research question rather than a given. The ACMG tiers place Tier 3 at 1/200 [10], but those thresholds come largely from populations where consanguinity is uncommon. Report your ranked gene list at your prespecified threshold and at plausible alternatives, so the decision’s sensitivity is visible.

How do I handle tribal or regional findings responsibly?

Agree the reporting granularity with your ethics committee before analysis, not after seeing results. Options include named groups, anonymised genomic clusters or national-level reporting only. Group-level findings affect people who never individually consented, and the non-discrimination provision applies directly [2].

Can my thesis recommend changes to the national screening panel?

No. It can produce ranked, criteria-scored evidence and discuss what that evidence would imply. Decisions on the national panel are made by ministerial resolution [2] and rest with health authorities, informed by counselling capacity, cost and implementation considerations beyond a doctoral study.

What if I cannot access national genomic data?

Name a fallback in the proposal. Options include a clinical exome or diagnostic cohort with its narrower sampling claim stated, a systematic review and meta-analysis of published regional autozygosity and carrier data, or a methodological study comparing ROH calling approaches on available data. Each is publishable and each de-risks the primary study.

Request a genetic-epidemiology topic feasibility and proposal review

Request a genetic-epidemiology topic feasibility and proposal review — PhD Assistance’s research-methodology team reviews your topic against this blueprint. You receive:

  • Novelty assessment: your question mapped against current UAE screening policy, national genomic work and the regional literature
  • Exposure-definition check: autozygosity versus reported consanguinity, and which your data and question support
  • ROH pipeline review: calling tool, parameter set, justification strategy and sensitivity analysis plan
  • Carrier-frequency analysis plan: classification rule, uncertain-variant handling, interval method and precision calculation
  • Panel-evidence framework: criteria operationalisation, scoring plan and threshold sensitivity design
  • Ethics and governance map: approvals, consent scope, reporting granularity and return-of-results pathway
  • Fallback design: a documented lower-risk alternative if data access is delayed

To start, share: your draft topic or research questions, your target university and programme, and what you currently know about data access and ethics approval. The review supports your own proposal writing; it does not replace it, and it does not provide clinical, counselling or reproductive guidance.

Related support: research proposal development · bioinformatics and biostatistics analysis · systematic literature review

Your next step

Want to apply this to your own study?

  1. 1Tell us your topic, data and deadline
  2. 2A subject mentor reviews your work with you
  3. 3You get a plan and a quote
Research design

References

  1. Ministry of Health and Prevention, United Arab Emirates. MoHAP announces mandatory genetic testing as part of premarital screening for Emiratis starting early January 2025. News release; 25 December 2024. gov.ae
  2. United Arab Emirates. Federal Decree-Law No. 49 of 2023 Regulating the Use of the Human Genome. Issued 2 October 2023; in force 1 December 2023. gov.ae
  3. Mousa M, Olbrich RM, Wohlers I, et al. The UAE Genome Program: unique genetic insights from 43,608 individuals. medRxiv [preprint, not peer reviewed]. 2025. doi:10.1101/2025.09.12.25334546
  4. Al-Gazali LI, Bener A, Abdulrazzaq YM, Micallef R, Al-Khayat AI, Gaber T. Consanguineous marriages in the United Arab Emirates. J Biosoc Sci. 1997;29(4):491–497. doi:10.1017/S0021932097004914
  5. Bizzari S, Nair P, Hana S, Deepthi A, Al-Ali MT, Al-Gazali L, El-Hayek S. Spectrum of genetic disorders and gene variants in the United Arab Emirates national population: insights from the CTGA database. Front Genet. 2023;14:1177204. doi:10.3389/fgene.2023.1177204
  6. Aamer W, Al-Maraghi A, Syed N, et al. Burden of Mendelian disorders in a large Middle Eastern biobank. Genome Med. 2024;16:46. doi:10.1186/s13073-024-01307-6
  7. Ceballos FC, Joshi PK, Clark DW, Ramsay M, Wilson JF. Runs of homozygosity: windows into population history and trait architecture. Nat Rev Genet. 2018;19(4):220–234. doi:10.1038/nrg.2017.109
  8. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405–424. doi:10.1038/gim.2015.30
  9. Newcombe RG. Two-sided confidence intervals for the single proportion: comparison of seven methods. Stat Med. 1998;17(8):857–872. doi:10.1002/(SICI)1097-0258(19980430)17:8<857::AID-SIM777>3.0.CO;2-E
  10. Gregg AR, Aarabi M, Klugman S, et al. Screening for autosomal recessive and X-linked conditions during pregnancy and preconception: a practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2021;23(10):1793–1806. doi:10.1038/s41436-021-01203-z
  11. Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81(3):559–575. doi:10.1086/519795
  12. Narasimhan V, Danecek P, Scally A, Xue Y, Tyler-Smith C, Durbin R. BCFtools/RoH: a hidden Markov model approach for detecting autozygosity from next-generation sequencing data. Bioinformatics. 2016;32(11):1749–1751. doi:10.1093/bioinformatics/btw044
  13. Szpiech ZA, Blant A, Pemberton TJ. GARLIC: Genomic Autozygosity Regions Likelihood-based Inference and Classification. Bioinformatics. 2017;33(13):2059–2062. doi:10.1093/bioinformatics/btx102
  14. Jain R, Bizzari S, Ramaswamy S, Hasham KF, et al. Pathogenic variation underlying rare diseases in an Arab population: implications for screening programs. Genet Med Open. 2025;3:103446. doi:10.1016/j.gimo.2025.103446
  15. Alblooshi K, Sharaf R, Shenbagam S, Sinha S, Jain R, et al. Citywide premarital genomic screening in a Middle Eastern population. Nat Med. 2026;32:1511–1518. doi:10.1038/s41591-026-04251-7

How to cite this article

Choose your referencing style, then copy. Check your university's or journal's guide for local variations.

PhD Assistance. (2026, October 5). Runs-of-Homozygosity Analysis for Recessive Disease Burden and Expanded Carrier Screening Panels in the UAE: A PhD Research Design Blueprint. PhD Research Lab. https://www.phdassistance.com/academy/how-to-article/roh-autozygosity-phd-research-design-uae/

PhD Assistance. “Runs-of-Homozygosity Analysis for Recessive Disease Burden and Expanded Carrier Screening Panels in the UAE: A PhD Research Design Blueprint.” PhD Research Lab, 5 Oct. 2026, www.phdassistance.com/academy/how-to-article/roh-autozygosity-phd-research-design-uae. Accessed [access date].

PhD Assistance (2026) Runs-of-Homozygosity Analysis for Recessive Disease Burden and Expanded Carrier Screening Panels in the UAE: A PhD Research Design Blueprint. PhD Research Lab. Available at: https://www.phdassistance.com/academy/how-to-article/roh-autozygosity-phd-research-design-uae/ (Accessed: [access date]).

PhD Assistance. “Runs-of-Homozygosity Analysis for Recessive Disease Burden and Expanded Carrier Screening Panels in the UAE: A PhD Research Design Blueprint.” PhD Research Lab. October 5, 2026. https://www.phdassistance.com/academy/how-to-article/roh-autozygosity-phd-research-design-uae.

PhD Assistance, ‘Runs-of-Homozygosity Analysis for Recessive Disease Burden and Expanded Carrier Screening Panels in the UAE: A PhD Research Design Blueprint’ (PhD Research Lab, 5 October 2026) <https://www.phdassistance.com/academy/how-to-article/roh-autozygosity-phd-research-design-uae> accessed [access date].

PhD Assistance. Runs-of-Homozygosity Analysis for Recessive Disease Burden and Expanded Carrier Screening Panels in the UAE: A PhD Research Design Blueprint [Internet]. PhD Research Lab; 2026 Oct 5 [cited [access date]]. Available from: https://www.phdassistance.com/academy/how-to-article/roh-autozygosity-phd-research-design-uae/

PhD Assistance, “Runs-of-Homozygosity Analysis for Recessive Disease Burden and Expanded Carrier Screening Panels in the UAE: A PhD Research Design Blueprint,” PhD Research Lab, Oct. 5, 2026. [Online]. Available: https://www.phdassistance.com/academy/how-to-article/roh-autozygosity-phd-research-design-uae. [Accessed: [access date]].

@online{roh_autozygosity_phd_research_design_uae_2026,
  author = {{PhD Assistance}},
  title = {{Runs-of-Homozygosity Analysis for Recessive Disease Burden and Expanded Carrier Screening Panels in the UAE: A PhD Research Design Blueprint}},
  organization = {PhD Research Lab},
  date = {2026-10-05},
  url = {https://www.phdassistance.com/academy/how-to-article/roh-autozygosity-phd-research-design-uae},
  urldate = {[access date]}
}

This article is for learning. Use it to understand the method and cite the original sources above in your own work; do not copy or submit it as your own.

Runs-of-Homozygosity Analysis for Recessive Disease Burden and Expanded Carrier Screening Panels in the UAE: A PhD Research Design BlueprintNeed help applying it to your study?
Research design