Free scoping read. A written assessment of your manuscript in two business days, at no cost.

Start one
Contents

Academy · Research Writing · oncology dissertation topics

Choosing an Oncology Dissertation Topic in Molecular and Cellular Cancer Biology

The binding constraint on a molecular oncology dissertation is almost never the interest of the question. It is supervision time and resources: lack of protected research time was the major barrier for 61 per cent of 130 responding supervisors, and only 11 per cent said they had all the resources their students’ projects needed. A topic chosen without testing against those two facts will be abandoned or rescued, and rescued projects produce thin dissertations.

Pubrica AcademyWritten by the medical writing and biostatistics teams Reviewed byPubrica biostatistics team Published9 October 2026 Last checked9 October 2026 Reading time13 minutes Sources35 peer-reviewed

Students choosing a dissertation topic in the molecular and cellular basis of cancer usually start from the biology, which is the right instinct and the wrong first step. The questions in this field are not in short supply — the hallmarks framework alone organises cancer biology into six core hallmarks, two enabling characteristics and two emerging hallmarks, and more than 90 approved small-molecule targeted drugs generate a resistance and selection literature that regenerates every year. What is in short supply is supervised time, validated reagents and a model system that can answer the question you picked. Topic selection is a feasibility exercise wearing the costume of an intellectual one.

What you will learn

  • How to test a topic against the supervision and resource constraints that actually determine outcomes
  • How to use the hallmarks framework as a map for locating a tractable question
  • Which classes of question in molecular oncology are dissertation-scale and which are not
  • Where the dissertation-scale questions sit in a landscape of more than 90 approved targeted agents
  • How to scope a topic so that a negative result is still a dissertation
  • How to convert a topic into a question with a measurable endpoint

What to unlearn

  • “Pick the most interesting question and the practicalities will follow.” They will not. Only 11 per cent of responding supervisors said they had all the resources their students’ projects needed, so for the large majority of students the resource position is a constraint on the question rather than a detail of its execution.
  • “My supervisor will have time for this.” Lack of protected research time was the major barrier reported by 61 per cent of 130 supervisors. Assume supervision is scarce and design a topic that can survive weeks without it.
  • “A heavily cited framework is a source of answers.” The hallmarks paper has been cited more than 59,000 times, which tells you how useful the organisation is and nothing about whether any individual hallmark contains a tractable question. Citation counts measure uptake, not available opportunity.
  • “Targeted therapy is a crowded field with nothing left for a student.” More than 90 approved small-molecule targeted agents generate exactly the opposite situation: a resistance, patient-selection and combination space that is larger than the trial system can explore.
  • “Basic science is the default for a biology dissertation.” In one study of student research projects, 22 per cent chose basic science against 61 per cent clinical and 10 per cent epidemiological. You will be in a minority, which has consequences for supervision availability and for who assesses you.

The constraint that actually decides the outcome

Two figures from a survey of research supervisors should shape topic selection more than any consideration of scientific interest. Of 130 supervisors who responded out of 463 invited, 61 per cent reported lack of protected research time as the major barrier, and only 11 per cent said they had all the resources their students’ projects needed 13,17.

Read those as statements about your working conditions rather than as complaints about the system. The first means supervision will be intermittent, so a topic whose every step requires a judgement call from your supervisor will stall 13. The second means that for roughly nine students in ten, something the project needs will be missing — a reagent, an instrument slot, a cell line, an animal licence, a software licence, a technician’s hours — and the time spent discovering which thing is missing is time taken from the dissertation.

The response is not to choose a dull topic. It is to run the resource test before committing, in writing, with the gaps named. A topic with one identified gap and a plan for it is a better bet than a topic with no identified gaps, because the second has not been examined. And note the response rate on that survey: 130 of 463 invited. Supervisors who are too overloaded to answer a survey are plausibly the most overloaded of all, which if anything makes the picture more rather than less severe.

Our page on quality in doctoral biomedical writing covers what examiners look for at the far end of this, and cancer research subject support is where the scientific scoping gets done.

61%of supervisors cited lack of protected research time as the major barrier
11%said they had all the resources their students’ projects needed
130supervisors responded, of 463 invited

The supervision and resource position a dissertation topic has to be chosen against, from a survey of research supervisors. Note the response rate: the most overloaded supervisors are also the least likely to have replied.

The hallmarks framework as a map, not an answer

The hallmarks framework organises cancer biology into six core hallmarks, two enabling characteristics and two emerging hallmarks, and the paper setting it out has been cited more than 59,000 times. That citation figure is worth interpreting carefully, because students read it as a signal of where the science is rather than of what the paper is for.

What the framework gives you is a map. It lets you locate a candidate question — this is a question about proliferative signalling, that one about immune evasion, this one about the enabling characteristics rather than a core hallmark — and it lets you see what you are not addressing. That second function is the more useful one for a dissertation, because the commonest topic failure is a question that straddles three hallmarks and therefore needs three sets of methods.

What the framework does not give you is a hypothesis. Naming a hallmark says where your question lives; it does not say what you expect to happen 32. A dissertation topic has to specify a mechanism precise enough to predict an observation, and the framework operates at a level of abstraction above that. Use it to position and to exclude, then do the mechanistic work separately.

One practical use of the map: locate your supervisor’s existing work on it first. A question in the same hallmark as the group’s current work inherits reagents, protocols, controls and — most importantly, given the supervision data — the ability of people around you to answer a question without your supervisor 13. A question two hallmarks away from the group’s expertise is a solo project with a part-time supervisor, which is the configuration that fails.

The hallmarks framework, and what it is and is not for

ComponentCountUse in topic selection
Core hallmarks6Locate the question; most dissertation topics should sit inside one
Enabling characteristics2Often where methods-adjacent and mechanism-enabling questions live
Emerging hallmarks2Less settled, so more open questions and less inherited protocol
Citations to the frameworkMore than 59,000Evidence of uptake as an organising scheme — not evidence of where opportunity lies

The hallmarks framework’s own structure and citation count. The framework organises cancer biology conceptually; it does not supply a mechanism, so naming a hallmark is not the same as stating a hypothesis.

Classes of question, and which are dissertation-scale

Most molecular oncology dissertation topics fall into a small number of classes, and they differ sharply in tractability.

Target validation in a defined context — does perturbing this gene or protein change this phenotype in this system? Tractable, because the endpoint is clear and the result is informative either way. This is the safest class for a time-limited project.

Resistance mechanism — why does a given agent stop working in a given setting? Tractable if a resistant model exists or can be generated within the timeline, which is the question to ask before committing. Generating resistant lines is itself months of work and is frequently underestimated.

Patient selection or biomarker association — which tumours respond, and what marks them? Tractable only if the specimens exist, with governance in place and sufficient numbers. The feasibility question is entirely about access, not about biology.

Combination rationale — why should two agents be used together? Tractable at the level of a mechanistic prediction tested in a model; not tractable as an empirical screen, because the combinatorial space is too large.

Novel target discovery — is there something here nobody has drugged? The least tractable class for a dissertation, because the discovery step has an unbounded timeline and a high probability of returning nothing usable.

Methods or measurement development — can we measure this better? Often an excellent dissertation, and systematically undervalued by students, who treat it as infrastructure rather than as a contribution. It has the advantage that the deliverable is in your control.

For the clinical and epidemiological alternatives to these, see epidemiology designs for clinical trials, and for how to write any of them up as a proposal, planning and writing a cancer research proposal.

Classes of molecular oncology dissertation topic, by tractability

Class of questionWhat makes it tractableWhat usually kills itDissertation-scale?
Target validation in a defined contextA clear endpoint and an informative result either wayA model lacking the pathway the question concernsYes — the safest class for a time-limited project
Resistance mechanismA resistant model already in handGenerating the resistant line, which is months of work on its ownYes, if the model exists; no, if it has to be built first
Patient selection or biomarkerSpecimens with governance in place and sufficient numbersAccess, accrual rate and specimen quality — not biologyOnly when access is confirmed before commitment
Combination rationaleA mechanistic prediction that one pairing should workAttempting it as an empirical screen across a space too large to screenYes, as a mechanistic test; no, as a screen
Novel target discoveryVery little, at dissertation timescaleAn unbounded discovery step with a high chance of returning nothing usableRarely
Methods or measurement developmentThe deliverable is in your control and stands whatever it showsBeing treated as infrastructure rather than written up as a contributionYes — systematically undervalued by students

Classification by tractability against dissertation timescales. The resource context is documented: of 130 responding supervisors, 61 per cent cited lack of protected research time as the major barrier and only 11 per cent said they had all the resources their students’ projects needed.

Where the questions are in a landscape of ninety agents

More than 90 small-molecule targeted drugs are approved for clinical use, against a global cancer burden projected to rise by roughly 50 per cent over twenty years 9. Students read the first figure as crowding and the second as urgency. Both readings are available, but the useful inference is different: ninety approved agents is a large installed base of interventions whose behaviour in specific contexts is only partly characterised, and that is where dissertation-scale questions live 9,21.

Three specific openings recur. Resistance to an approved agent in a defined subtype is frequently unresolved even for drugs that have been in clinic for years, because the clinical trial that registered the agent was not designed to characterise relapse. Determinants of response within a licensed indication — the subset question — is open for most agents and has direct clinical consequence 21. Mechanistic grounds for a specific combination matters precisely because ninety agents generate a combination space that cannot be explored empirically in trials; a principled reason to test one pair rather than another is a real contribution.

What to avoid is a topic premised on the field being empty. A dissertation claiming that nobody has considered a target is nearly always wrong, and the usual explanation for an apparently neglected target is that someone tried and it did not work. If your topic involves such a target, find out what happened and build the project around what was different — a new chemotype, a different context, a better readout. That is a stronger dissertation than an unexamined novelty claim, and it is a question your examiners will respect. Oncology and haematology therapeutic support covers the landscape reading this requires.

Scoping to the model and the clock

Scope is where good topics become undeliverable ones. The discipline is to work backwards from the model system and the time available, rather than forwards from the question.

Start with the model. Ask what the model system available to you can actually show. A cell line lacking an intact pathway cannot answer a question about that pathway; a xenograft without an immune compartment cannot answer an immune question; a two-dimensional culture cannot answer a question about architecture or gradient. This sounds obvious and is the single commonest reason a dissertation produces uninterpretable data: the question and the model were chosen separately.

Then the clock. Count backwards from the submission date through writing, analysis, repeat experiments, the first full experiment, optimisation, reagent lead times and training. Optimisation and reagent lead times are the two that students omit and that reliably consume months. Whatever remains is the window for the experiment that answers the question, and if the window is a few weeks, the question needs to be smaller.

Then the endpoint. Name the measurement, and check that it is validated in your hands, in your matrix and at your scale. An unvalidated readout contaminates every number the project produces, and it is the feasibility claim students most often assume. If validating the readout is itself a month of work, that month belongs in the plan.

A topic scoped this way often looks modest on paper. It also finishes, and a finished modest project is a better dissertation than an abandoned ambitious one. Our page on structuring a research proposal around thrusts and tasks covers the decomposition that makes this concrete.

Choosing basic science, when most students do not

In a study of student research projects in which 358 of 652 eligible students returned both questionnaires, 61 per cent undertook clinical projects, 22 per cent basic science, 10 per cent epidemiological and 7 per cent other 13,17. If you are proposing a molecular or cellular cancer biology project, you are in the 22 per cent, and that has three practical consequences.

Supervision in your area may be thinner on the ground, which interacts badly with the protected-time problem 13,17. Check not only that your supervisor works in the field but that someone else in the building does, because that is who will answer the question that cannot wait a fortnight.

Your assessors may be more accustomed to clinical projects. Write the mechanism so that a competent non-specialist can follow it, and state explicitly what the result would mean for anyone outside your subfield — not as inflation, but as translation.

And the failure modes differ. Clinical and epidemiological projects fail on recruitment and data access; basic science projects fail on reagent availability, model limitations and unvalidated readouts 13. Diagnose the failure modes of your own category rather than borrowing a generic risk list. The resource figures reinforce this: with only 11 per cent of supervisors reporting that they had all the resources their students’ projects needed, a bench project should assume something will be missing and identify it early.

22%of students chose basic science projects
61%chose clinical projects
10%chose epidemiological projects

Project type among students in a study where 358 of 652 eligible students returned both questionnaires; a further 7 per cent chose other project types. A molecular oncology dissertation puts you in the smaller group, with consequences for supervision availability and for who assesses you.

Designing a topic that survives a negative result

The most important property of a dissertation topic is that both answers are publishable as a dissertation. A topic that yields a thesis only if the hypothesis is confirmed is a gamble, and the student carries the risk.

Three design moves achieve this. Ask a discriminating question rather than a confirmatory one. “Does X contribute to Y?” has a useless negative answer; “is the limiting step X or Z?” has two informative answers, because ruling one out is a result. Make the measurement the contribution where you can. If part of the dissertation is establishing a readout or a model, that work stands regardless of what the readout shows. Pre-specify the interpretation of a null. Write down, before running the experiment, what a null result would mean and what it would rule out — and check that the experiment is sensitive enough for the null to mean anything. An underpowered null means nothing, which is the trap.

Discuss this explicitly with your supervisor at topic selection, because it is the conversation that most reliably improves a project and the one least likely to happen spontaneously. The question to ask is simple: if the central experiment comes out negative, what is the dissertation? If neither of you can answer, the topic needs reshaping before any work starts.

From topic to question to title

A topic is a region; a question is a sentence with an answer. The conversion is the work, and it is worth doing on paper before the first experiment.

Write the question as a single sentence containing the system, the perturbation, the comparator and the measured endpoint 1,15. If any of the four is missing, the question is still a topic. Then write the two headline sentences you would be able to write at the end — one for each possible answer. If either of them is a sentence you would not want to write, the question is wrong. This takes twenty minutes and saves months.

Then the title. Dissertation titles in this field default to a form that says nothing: investigating the role of a molecule in a cancer. It names a region and a technique and withholds the finding, which is precisely backwards from what a title should do. A title that names the system, the relationship and the specificity — even provisionally, to be revised when the result is in — forces the question to be specific and gives your examiners something to hold.

Finally, write the one-paragraph rationale: why this question, why now, why this system, why this is the next thing to know. Our page on writing the rationale covers how to build that paragraph so that it argues rather than asserts, and thesis and dissertation support is where the writing-up stage gets handled.

Turning an oncology dissertation topic into a question

A region of biology with a technique attached

Investigating the role of autophagy in chemoresistance in triple-negative breast cancer. This project will use a panel of cell lines to characterise autophagic flux following treatment and explore the relationship between autophagy and resistance, with the aim of identifying novel therapeutic targets.

“Investigating the role of” names a region, not a question. There is no comparator and no endpoint, so no result can be stated as an answer.

“Characterise” and “explore” are activities. They can be performed indefinitely and completed never, which is how a dissertation runs out of time with data but no finding.

“A panel of cell lines” does not say whether the models can show the effect. Resistance that is intrinsic rather than acquired may not be the phenomenon the question is about.

No endpoint and no assay validation. An unvalidated readout contaminates every number the project will produce, and it is the feasibility claim students most often assume.

“Identifying novel therapeutic targets” is an aspiration three steps beyond the work. It also hides the fact that no single experiment in the project can be said to have succeeded or failed.

A negative result leaves nothing. If autophagic flux turns out not to change, there is no dissertation, and the student carries that risk.

A discriminating question with both answers useful

Does inhibition of autophagy restore sensitivity to this agent in acquired-resistant cells, or is resistance in this setting independent of autophagy? Two accounts appear in the literature for this tumour type and they predict opposite results from the same manipulation, which has not been performed in a line with acquired rather than intrinsic resistance. We hold two acquired-resistant derivatives generated in the group in [year], with the parental line as comparator; the endpoint is a dose-response shift measured by the assay we report in Chapter 2, for which we establish the dynamic range and the inter-day coefficient of variation as the first piece of work. A positive result identifies a resensitisation strategy and specifies which resistance route it applies to. A negative result excludes autophagy as the operative route in this setting and redirects the remaining work to the bypass candidate, which the same cells and the same assay can address without new reagents. Either way the assay characterisation stands as a contribution. The question sits within one hallmark; we do not address the immune or metabolic dimensions, which are outside the group’s reagent set.

The question is discriminating rather than confirmatory, so ruling one account out is itself a result.

The models are named and the distinction that matters — acquired against intrinsic resistance — is made explicit, so the system can actually address the question.

The endpoint is specified and its characterisation is built into the plan as the first piece of work, which removes the commonest source of uninterpretable bench data.

Both outcomes are written out in advance, and the negative one redirects the remaining time using reagents already in hand.

The assay work is claimed as a contribution in its own right, which is the cheapest available insurance against a null central result.

The question is scoped to one hallmark and the exclusions are stated, so the project does not require three sets of methods.

With lack of protected supervision time reported by 61 per cent of supervisors and only 11 per cent fully resourced, a topic that needs a judgement call at every step and only works if the hypothesis holds is the configuration that fails. The second version can be executed during a fortnight when nobody is available to advise.

Take-home message

Take-home message

Choose an oncology dissertation topic against the supervision and resource position rather than against interest: lack of protected research time was the major barrier for 61 per cent of 130 responding supervisors, and only 11 per cent said they had all the resources their students’ projects needed. Use the hallmarks framework — six core hallmarks, two enabling characteristics and two emerging hallmarks — to locate the question and, more usefully, to exclude the dimensions you will not address; its citation count of more than 59,000 measures uptake, not opportunity. In a landscape of more than 90 approved targeted agents the dissertation-scale questions are in resistance, patient selection and combination rationale rather than in novel target discovery. Scope backwards from the model system, the clock and the validated endpoint, and design the question so that both answers are a dissertation. Remember you are in the 22 per cent of students doing basic science rather than the 61 per cent doing clinical work, with the supervision and assessment consequences that implies.

How to apply this to your study

How to apply this to your own study

  1. Run the resource test in writing before committing. List every reagent, instrument slot, cell line, licence and technician hour the project needs, mark what you do not have, and get each gap confirmed or closed before the first experiment.
  2. Locate your supervisor’s current work on the hallmarks map, then pick nearby. A question in the same hallmark inherits protocols, controls and people who can answer you when your supervisor cannot.
  3. Check the model before the question. Ask explicitly what your available system cannot show, and discard any question that needs a feature the model lacks.
  4. Count the clock backwards. Writing, analysis, repeats, first full experiment, optimisation, reagent lead times, training. Optimisation and lead times are the two students omit, and they consume months.
  5. Write both headline sentences. One for each possible answer. If either is a sentence you would not want to write, the question is wrong.
  6. Build the endpoint characterisation into the plan as work. If validating the readout takes a month, that month is in the timeline — and the characterisation is a contribution that stands whatever the central result shows.
  7. Ask your supervisor the null question explicitly. If the central experiment comes out negative, what is the dissertation? Reshape the topic until there is an answer.
  8. Replace an “investigating the role of” title. Name the system, the relationship and the specificity, even provisionally. A specific title forces a specific question.

Common questions

Choose against your constraints first and your interests second, because the constraints decide the outcome. Lack of protected research time was the major barrier reported by 61 per cent of 130 responding supervisors, and only 11 per cent said they had all the resources their students’ projects needed — so assume supervision will be intermittent and that something the project needs will be missing. Practically: locate your supervisor’s existing work on the hallmarks map and choose a question inside the same hallmark, so you inherit reagents, protocols and colleagues who can answer a question when your supervisor is unavailable. Then test the question against the model you actually have, the time you actually have, and whether the endpoint can be measured reliably in your hands.

As a map, yes; as a source of questions, no. The framework organises cancer biology into six core hallmarks, two enabling characteristics and two emerging hallmarks, and the paper setting it out has been cited more than 59,000 times — which tells you how widely the organisation has been adopted, not where opportunity lies. Its two genuine uses in topic selection are locating a candidate question and, more valuably, excluding the dimensions you will not address; the commonest topic failure is a question straddling three hallmarks and therefore needing three sets of methods. What the framework cannot do is supply a mechanism. Naming a hallmark says where your question lives; a dissertation needs a mechanism specific enough to predict an observation.

Yes, and the apparent crowding is the reason. More than 90 small-molecule targeted drugs are approved for clinical use, which is a large installed base of interventions whose behaviour in specific contexts is only partly characterised. Three openings recur at dissertation scale. Resistance to an approved agent in a defined subtype is often unresolved even for long-established drugs, because the registration trial was not designed to characterise relapse. Determinants of response within a licensed indication — the subset question — is open for most agents with direct clinical consequence. And mechanistic grounds for a specific combination matter precisely because ninety agents generate a combination space trials cannot explore. What to avoid is a topic premised on nobody having considered a target; usually someone tried.

Needing more than one set of methods, more than one model system, or a step with an unbounded timeline. Three specific warning signs. The question spans more than one hallmark, which means you will have to establish competence in two areas and will do neither well. The plan includes generating a model — a resistant line, a knock-in, a new xenograft — before the real experiment starts, because model generation is routinely underestimated and can consume the whole project. Or the first step is a discovery or screen whose output you cannot predict, which makes everything downstream conditional. Scope backwards instead: start from what your model can show, subtract writing, analysis, repeats, optimisation and reagent lead times from the clock, and size the question to what remains.

Design for it from the start, because a topic that only yields a thesis when the hypothesis holds puts the risk on you. Three moves. Ask a discriminating question rather than a confirmatory one: does X contribute to Y has a useless negative answer, whereas is the limiting step X or Z has two informative answers. Make part of the work a measurement or model contribution that stands regardless — if you characterise an assay or establish a system, that is a chapter whatever it then shows. And pre-specify what a null would mean and what it would rule out, then check that the experiment is sensitive enough for the null to mean anything, since an underpowered null means nothing. Ask your supervisor directly: if the central experiment is negative, what is the dissertation?

Either, but know which minority you are joining and what it costs. In one study where 358 of 652 eligible students returned both questionnaires, 61 per cent undertook clinical projects, 22 per cent basic science, 10 per cent epidemiological and 7 per cent other. Choosing basic science has three consequences. Supervision in your specific area may be thinner, which compounds the protected-time problem — so check that someone besides your supervisor in the building works on this. Your assessors may be more used to clinical projects, so the mechanism has to be legible to a competent non-specialist. And the failure modes differ: clinical projects fail on recruitment and data access, bench projects on reagent availability, model limitations and unvalidated readouts. Diagnose your own category’s risks rather than a generic list.

Five questions, and the answers matter more than the topic. What does this project need that the group does not currently have, and who closes that gap? What is the realistic amount of time you will have for this, given that lack of protected research time was the major barrier for 61 per cent of surveyed supervisors? If the central experiment comes out negative, what is the dissertation? Who else here can answer a technical question when you are unavailable? And has anyone, here or elsewhere, already tried this — and what happened? That last question is the one students skip and the one most likely to redirect a topic productively, because an apparently neglected question usually has a history.

Not as investigating the role of a molecule in a cancer, which is the default and which names a region and a technique while withholding any finding. A title should do the opposite: name the system, the relationship and the specificity, even provisionally, with the expectation of revising it once the result is in. The practical benefit is upstream rather than cosmetic — a specific title cannot be written unless the question is specific, so drafting the title early is a test of whether you have converted a topic into a question. Write the question as a single sentence containing the system, the perturbation, the comparator and the measured endpoint; if any of the four is missing, you still have a topic.

References and sources

  1. Willis LD. Formulating the Research Question and Framing the Hypothesis. Respiratory Care. 2024;69(4). doi:10.4187/respcare.10975 read 9 October 2026. doi.org/10.4187/respcare.10975
  2. Patil SG. Budgeting the Project: Detailing the Required Funding and Expenditure. Grant writing for medical and healthcare professionals (Parija SC, Kate V, eds), Springer, Singapore. 2023:77-91. doi:10.1007/978-981-19-7018-4_7 read 9 October 2026. doi.org/10.1007/978-981-19-7018-4_7
  3. Rajappa M. Checklist for Grant Proposal: Mandatory Elements. Grant writing for medical and healthcare professionals (Parija SC, Kate V, eds), Springer, Singapore. 2023:161-165. doi:10.1007/978-981-19-7018-4_14 read 9 October 2026. doi.org/10.1007/978-981-19-7018-4_14
  4. Joshi M, Joshi P, Smith JA. Dealing with the Rejected Grant Proposal: Learning from the Mistakes. Grant writing for medical and healthcare professionals (Parija SC, Kate V, eds), Springer, Singapore. 2023:179-187. doi:10.1007/978-981-19-7018-4_16 read 9 October 2026. doi.org/10.1007/978-981-19-7018-4_16
  5. Rahi M, Sharma S, Yadav B. Funders' Perspective: What the Funding Body Expects from Researchers? Grant writing for medical and healthcare professionals (Parija SC, Kate V, eds), Springer, Singapore. 2023:21-31. doi:10.1007/978-981-19-7018-4_3 read 9 October 2026. doi.org/10.1007/978-981-19-7018-4_3
  6. Dubashi B, Ramadass B. Funding Organizations for Science and Health Care. Grant writing for medical and healthcare professionals (Parija SC, Kate V, eds), Springer, Singapore. 2023:33-52. doi:10.1007/978-981-19-7018-4_4 read 9 October 2026. doi.org/10.1007/978-981-19-7018-4_4
  7. Parija SC, Kate V (eds). Grant writing for medical and healthcare professionals. Springer, Singapore. 2023. doi:10.1007/978-981-19-7018-4 read 9 October 2026. doi.org/10.1007/978-981-19-7018-4
  8. Weidmann AE, Cadogan CA, Fialova D, et al. How to write a successful grant application: guidance provided by the European Society of Clinical Pharmacy. International Journal of Clinical Pharmacy. 2023;45(3):781-786. doi:10.1007/s11096-023-01543-7 read 9 October 2026. doi.org/10.1007/s11096-023-01543-7
  9. Han C, Zhan Q. Precision medicine revolutionizes cancer diagnosis and treatment. Medical Review. 2023;2(6):541-543. doi:10.1515/mr-2022-0041 read 9 October 2026. doi.org/10.1515/mr-2022-0041
  10. Nachiappa Ganesh R. Tips for Improvising the Chance of Getting the Grant. Grant writing for medical and healthcare professionals (Parija SC, Kate V, eds), Springer, Singapore. 2023:167-177. doi:10.1007/978-981-19-7018-4_15 read 9 October 2026. doi.org/10.1007/978-981-19-7018-4_15
  11. Parija SC, Shrivastava SR. What Is Your Aim (Ambition-Interest-Mission), and What You Want to Do? Grant writing for medical and healthcare professionals (Parija SC, Kate V, eds), Springer, Singapore. 2023:3-12. doi:10.1007/978-981-19-7018-4_1 read 9 October 2026. doi.org/10.1007/978-981-19-7018-4_1
  12. Kate V, Gupta D, Balakrishnan G. Writing a Grant Proposal for a Single Centre Study: Step-by-Step Approach. Grant writing for medical and healthcare professionals (Parija SC, Kate V, eds), Springer, Singapore. 2023:63-76. doi:10.1007/978-981-19-7018-4_6 read 9 October 2026. doi.org/10.1007/978-981-19-7018-4_6
  13. Hart J, Hakim J, Kaur R, et al. Research supervisors' views of barriers and enablers for research projects undertaken by medical students; a mixed methods evaluation of a post-graduate medical degree research project program. BMC Medical Education. 2022;22:370. doi:10.1186/s12909-022-03429-0 read 9 October 2026. doi.org/10.1186/s12909-022-03429-0
  14. Monavarian M. Basics of scientific and technical writing: Grant proposals. MRS Bulletin. 2021;46(5):455-457. doi:10.1557/s43577-021-00105-4 read 9 October 2026. doi.org/10.1557/s43577-021-00105-4
  15. Dhir SK, Gupta P. Formulation of Research Question and Composing Study Outcomes and Objectives. Indian Pediatrics. 2021;58:584-588. read 9 October 2026. indianpediatrics.net
  16. Kan P, Mokin M, Mack WJ, et al. Strategies for writing a successful National Institutes of Health grant proposal for the early-career neurointerventionalist. Journal of NeuroInterventional Surgery. 2020. doi:10.1136/neurintsurg-2020-016964 read 9 October 2026. doi.org/10.1136/neurintsurg-2020-016964
  17. Moller R, Shoshan M. Does reality meet expectations? An analysis of medical students' expectations and perceived learning during mandatory research projects. BMC Medical Education. 2019;19:93. doi:10.1186/s12909-019-1526-x read 9 October 2026. doi.org/10.1186/s12909-019-1526-x
  18. Taylor RB. How to Write a Research Protocol. Medical Writing (Springer, Cham). 2018:239-260. doi:10.1007/978-3-319-70126-4_9 read 9 October 2026. doi.org/10.1007/978-3-319-70126-4_9
  19. Shepherd J, Frampton GK, Pickett K, et al. Peer review of health research funding proposals: A systematic map and systematic review of innovations for effectiveness and efficiency. PLoS ONE. 2018;13(5):e0196914. doi:10.1371/journal.pone.0196914 read 9 October 2026. doi.org/10.1371/journal.pone.0196914
  20. Boyack KW, Smith C, Klavans R. Toward predicting research proposal success. Scientometrics. 2018;114(2):449-461. doi:10.1007/s11192-017-2609-2 read 9 October 2026. doi.org/10.1007/s11192-017-2609-2
  21. Chae YK, Pan AP, Davis AA, et al. Path toward precision oncology: Review of targeted therapy studies and tools to aid in defining 'actionability' of a molecular lesion and patient management support. Molecular Cancer Therapeutics. 2017;16(12):2645-2655. doi:10.1158/1535-7163.MCT-17-0597 read 9 October 2026. doi.org/10.1158/1535-7163.MCT-17-0597
  22. Dey SK, Gerhardt K. How to Write a Competitive R01 Grant. Springer, Cham. 2016. doi:10.1007/978-3-319-41360-0 read 9 October 2026. doi.org/10.1007/978-3-319-41360-0
  23. Introduction: Fear of Uncertainty - Enter at Your Own Risk. How to Write a Competitive R01 Grant (Springer, Cham). 2016:1-18. doi:10.1007/978-3-319-41360-0_1 read 9 October 2026. doi.org/10.1007/978-3-319-41360-0_1
  24. Agency for Healthcare Research and Quality. Common Problems Identified During Peer Review. Agency for Healthcare Research and Quality (AHRQ). 2014. read 9 October 2026. ahrq.gov
  25. Tambe J, Ze Minkande J, Moifo B, et al. Students' perspectives on research and assessment of a model template designed to guide beginners in research in a medical school in Cameroon. BMC Medical Education. 2014;14:269. doi:10.1186/s12909-014-0269-y read 9 October 2026. doi.org/10.1186/s12909-014-0269-y
  26. Herbert DL, Barnett AG, Clarke P, Graves N. On the time spent preparing grant proposals: an observational study of Australian researchers. BMJ Open. 2013;3(5):e002800. doi:10.1136/bmjopen-2013-002800 read 9 October 2026. doi.org/10.1136/bmjopen-2013-002800
  27. Chan AW, Tetzlaff JM, Altman DG, et al. SPIRIT 2013 Statement: Defining Standard Protocol Items for Clinical Trials. Annals of Internal Medicine. 2013;158(3):200-207. doi:10.7326/0003-4819-158-3-201302050-00583 read 9 October 2026. doi.org/10.7326/0003-4819-158-3-201302050-00583
  28. Abdoul H, Perrey C, Amiel P, et al. Peer Review of Grant Applications: Criteria Used and Qualitative Study of Reviewer Practices. PLoS ONE. 2012;7(9):e46054. doi:10.1371/journal.pone.0046054 read 9 October 2026. doi.org/10.1371/journal.pone.0046054
  29. National Institute of Allergy and Infectious Diseases (NIAID). Strategy for NIH Funding: Ten Steps to a Winning R01 Application. National Institutes of Health (NIAID). 2012. read 9 October 2026. ipr.osu.edu
  30. Proctor EK, Powell BJ, Baumann AA, et al. Writing implementation research grant proposals: ten key ingredients. Implementation Science. 2012;7(1):96. doi:10.1186/1748-5908-7-96 read 9 October 2026. doi.org/10.1186/1748-5908-7-96
  31. Hollins Martin CJ, Fleming V. A 15-step model for writing a research proposal. British Journal of Midwifery. 2011;18:791-798. read 9 October 2026. napier-repository.worktribe.com
  32. Hanahan D, Weinberg RA. Hallmarks of Cancer: The Next Generation. Cell. 2011;144. doi:10.1016/j.cell.2011.02.013 read 9 October 2026. doi.org/10.1016/j.cell.2011.02.013
  33. Eve J. Writing a research proposal: planning and communicating your research ideas effectively. Library and Information Research. 2009;32(102):18-28. doi:10.29173/lirg92 read 9 October 2026. doi.org/10.29173/lirg92
  34. SUNY Geneseo Sponsored Research (adapted from Locke, Spirduso and Silverman, Proposals That Work, 2nd ed., Sage, 1987). Common Reasons Grant Proposals Are Rejected. SUNY Geneseo Office of Sponsored Research. read 9 October 2026. geneseo.edu
  35. National Institute of Mental Health. Grant Writing Assistance. National Institutes of Health (NIMH). read 9 October 2026. nimh.nih.gov

Every source above is a peer-reviewed article, a publisher’s record of one, or a standard issued by the body that owns it. Each carries the date it was last read. Where a standard moves, this article is revised rather than left up with a note.

How to cite this article

Pick your referencing style and copy it. Your university or target journal may vary the punctuation — check its guide if it has one.

Pubrica Academy. (2026, October 9). Choosing an Oncology Dissertation Topic in Molecular and Cellular Cancer Biology. Pubrica. https://pubrica.com/academy/research/dissertation-topics-covered-in-the-molecular-and-cellular-basics-of-cancer-and-development-of-target-specific-therapeutics-cancer-oncology/

This article is written to be learned from. Cite it if you are referring to what it says; if you are using the evidence behind it, cite the sources listed above instead.

Topic feasibility screen for a molecular oncology dissertation

Run this before committing to a topic, and run it again with your supervisor. Any section you cannot complete is a risk you are carrying unexamined. The screen is deliberately unromantic: it is designed to stop a project that would otherwise be abandoned or rescued at month nine.

MOLECULAR ONCOLOGY DISSERTATION TOPIC FEASIBILITY SCREEN

SECTION 1. THE QUESTION  (one sentence, four components)

  System: ______________________________
  Perturbation or exposure: ______________________________
  Comparator: ______________________________
  Measured endpoint: ______________________________

  Question as one sentence:
    ____________________________________________

  [ ] All four components present
     (if any is missing, this is still a topic, not a question)

  Is the question DISCRIMINATING or CONFIRMATORY?
    [ ] Discriminating - it asks which of two accounts holds
    [ ] Confirmatory - it asks whether one account holds
        (if confirmatory, see Section 6 before proceeding)


SECTION 2. POSITION ON THE MAP

  Hallmark(s) the question occupies: ______________________
  Hallmarks DELIBERATELY not addressed: ______________________
  [ ] The question sits inside ONE hallmark
     (if it spans more than one, list the method sets required:
      ______________________________  - can you learn all of them?)

  Where does my supervisor's current work sit on this map?
    ____________________________________________
  Distance from my question:  same hallmark / adjacent / far
  [ ] Same or adjacent  (far = a solo project with a part-time
      supervisor, which is the configuration that fails)

  Class of question  (tick one):
    [ ] Target validation in a defined context     - safest
    [ ] Resistance mechanism                       - safe IF model exists
    [ ] Patient selection or biomarker             - access decides it
    [ ] Combination rationale                      - as a test, not a screen
    [ ] Novel target discovery                     - rarely feasible
    [ ] Methods or measurement development         - undervalued, often best


SECTION 3. MODEL SYSTEM  (what it CANNOT show)

  Model available: ______________________________
  Does it retain the feature my question concerns?  yes / no / unsure
  What this model CANNOT show: ______________________________
  Does any of that overlap my question?  yes / no
     [ ] If yes, the question or the model must change

  If the model must be GENERATED (resistant line, knock-in, xenograft):
     Estimated time to generate: ______ months
     Has this group generated one before?  yes / no
     [ ] This time is in the timeline at Section 5


SECTION 4. RESOURCE TEST  (write the gaps down)

  Requirement                   Have it?  If not, who closes it?  By when?
  Cell lines or models           ____      __________________     ______
  Antibodies or probes           ____      __________________     ______
  Plasmids or constructs         ____      __________________     ______
  Instrument access and slots    ____      __________________     ______
  Animal licence or approval     ____      __________________     ______
  Human tissue governance        ____      __________________     ______
  Software or analysis licence    ____      __________________     ______
  Technician hours               ____      __________________     ______
  Consumables budget             ____      __________________     ______
  Training I still need          ____      __________________     ______

  Number of gaps: ______
  [ ] Every gap has a named owner and a date
     (benchmark for why this matters: only 11 per cent of surveyed
      supervisors said they had all the resources their students'
      projects needed)


SECTION 5. THE CLOCK  (count BACKWARDS from submission)

  Submission date .................................. ______________
  Writing up ....................................... ______ weeks
  Final analysis ................................... ______ weeks
  Repeat experiments (assume you will need them) .... ______ weeks
  The experiment that answers the question .......... ______ weeks
  Optimisation  [most underestimated] ............... ______ weeks
  Reagent and approval lead times ................... ______ weeks
  Training and competence ........................... ______ weeks
  Model generation, if any .......................... ______ weeks
  TOTAL ............................................. ______ weeks
  Weeks actually available .......................... ______

  [ ] Total is less than available
  If not, what gets cut: ______________________________


SECTION 6. THE ENDPOINT  (the readout that carries everything)

  Assay or measurement: ______________________________
  Validated in MY hands?  yes / no
  Validated in THIS matrix?  yes / no
  Validated at THIS scale?  yes / no
  Dynamic range known?  yes / no
  Inter-day variation known?  yes / no

  [ ] If any answer is no, characterising the endpoint is the FIRST
      piece of work and appears in the timeline
  [ ] That characterisation is claimed as a contribution in its own
      right  (this is the cheapest insurance against a null result)


SECTION 7. BOTH ANSWERS  (write them now, before any work)

  If the result is POSITIVE, my headline sentence is:
    ____________________________________________
    Would I want to write that sentence?  yes / no

  If the result is NEGATIVE, my headline sentence is:
    ____________________________________________
    Would I want to write that sentence?  yes / no
    What does the negative rule OUT: ______________________
    Is the experiment sensitive enough for that null to mean
    anything?  yes / no / unsure
       (an underpowered null means nothing)
    What the remaining time is redirected to: ______________
    Can that redirection use reagents already in hand?  yes / no

  [ ] Both answers produce a dissertation


SECTION 8. THE HISTORY QUESTION  (ask it explicitly)

  Has anyone tried this, here or elsewhere?  yes / no / unknown
  If yes, what happened: ______________________________
  What is different now: ______________________________
  [ ] I asked my supervisor this directly
  [ ] I searched for the failed or abandoned attempts, not only the
      published successes


SECTION 9. SUPERVISION REALITY CHECK

  Realistic supervisor contact: ______ per ______________
  Who else in the building can answer a technical question: ______
  [ ] The project can survive a fortnight with no supervisor input
  [ ] There is at least one other person here working in this area


SECTION 10. TITLE AND RATIONALE

  Provisional title, naming system, relationship and specificity:
    ____________________________________________
  [ ] It is NOT of the form "investigating the role of X in Y"

  One-paragraph rationale:
    Why this question: ______________________________
    Why now: ______________________________
    Why this system: ______________________________
    Why this is the next thing to know: ______________________

Pubrica AcademyWritten and maintained by the subject matter experts of Pubrica who do this work. Articles are revised when the standard underneath them moves, and each one carries the date it was last checked.

How Pubrica helps

Pubrica helps students and supervisors scope an oncology dissertation topic before the work starts: positioning the question against the hallmarks framework, testing it against the model system and the clock, and reshaping confirmatory questions into discriminating ones so that a null result is still a dissertation. We run the resource and endpoint checks that bench projects fail on, and we are candid when a topic needs to be smaller than the student wants it to be.

Find something else

Search the rest of the Academy.

Open with no account. Type what you are working on, or open the full library to filter by category, sub-topic or resource type.

Send us your proposed topic and the model system you have access to, and we will tell you whether it is a dissertation or a programme of work.

A scoping read costs nothing and takes two business days. You get a written view of what the work needs and who would hold it.

Get a free quote

Regulations

Check your project against this rule

  1. 1Your work
  2. 2Your details
  3. 3Send
What do you need?

Material you send is held under signed confidentiality, with access limited to the assigned team and encrypted transfer. It is not used to train anything and is not passed to third-party tools. Ask for the agreement before you send anything you would rather cover first.