Research gap finder: gaps in your field from real papers
See possible research gaps in the papers on your topic.
Free signals here, ideas with your Twin.
A few words, like graph neural networks for drug discovery.
Themes rarely studied together and themes growing fastest. Signals to check, not verdicts.
Sign in and your Twin turns a gap into ideas, then a proposal.
Your research topic
Graph neural networks for drug discovery
Find gapsCounted from the papers in our data.
What is a research gap?
A research gap is a question in your field that published research has not yet answered well. It is the space a thesis, a PhD or a research proposal sets out to fill. Most gaps are one of five types.
- Evidence gap. Studies on the same question disagree or contradict each other.
- Knowledge gap. Little or no research exists on the topic yet.
- Practical gap. What professionals do differs from what research recommends.
- Methodological gap. The question has been studied with only one kind of method.
- Population gap. A group, country or age range has been left out of the studies.
To find one, read the most recent papers on your topic and note the limits and future work their authors name. Then look for themes growing fast, and for common themes rarely studied together. ApplyKite's free finder counts both signals in the papers closest to your topic, from the published work of more than 820,000 professors.
Turn a gap into research ideas.
Your Twin reads the papers with you and suggests ideas, each naming its sources.
Uncertainty-aware toxicity prediction
Few of these papers report how sure their toxicity model is. [2] [5]
Protein language models for small datasets
Most of these papers train on large public sets. [1] [4]
One graph model across target families
Few of these papers test a model on more than one family. [3] [7]
Refine one idea into a proposal.
Sharpen the idea version by version, then your Twin writes the proposal with real citations.
Uncertainty-aware toxicity prediction
Version 2, in your words
Graph models that say how sure they are of a toxicity call, on small datasets.
The papers behind it
- [2] Graph neural networks for toxicity prediction. 2024.
- [5] Calibrated uncertainty in molecular property models. 2023.
From a gap to your research idea.
Questions
What is a research gap finder?
The research gap finder shows where your field is moving, from real recent papers. Type a topic and see, free, up to ten recent papers, the themes growing fastest and the pairs of themes rarely studied together. Sign in and ApplyKite Twin turns those gaps into research ideas, then helps you refine one into a proposal.
What is a research gap?
A research gap is a question your field has not answered well yet: a topic with little work on it, a method not yet tried on a problem, or results that disagree. A PhD or a research master's is usually built on one.
How do I find a research gap?
Read the most recent papers on your topic, note what their authors say is still missing, and look for themes that are growing fast or that are rarely studied together. This page counts those signals for you from real papers, so you know where to read first.
Is this based on real papers?
Yes. Every paper shown is a real publication with its title and year, and the signals are counts over real papers, not guesses. Nothing on this page is written by AI.
Is the research gap finder free?
Yes, with no sign-up. Sign in, free to start, if you want ApplyKite Twin to turn the gaps into research ideas grounded in those papers and help you shape one into a proposal.
What do rising themes and rarely-together themes mean?
A rising theme has many more papers recently than before, so the field is moving towards it. A rarely-together pair is two themes that are each common in your topic but seldom appear in the same paper, which can point to an open question.
How specific should my topic be?
A few words to a sentence works best, such as 'graph neural networks for drug discovery'. Too broad and the signals blur together; too narrow and there may be too few papers to count.
Read the guide: How to find a research gap →