Scientific communication is broken.
Progress in biomedical science is all about incrementally building off of the work of others. Currently, it takes far too long for one scientist's work to reach all the other scientists that would benefit. Mark2Cure will make scientific communication more efficient for everyone, leading to faster discoveries and cures for disease.
To be successful, we need your help. Mark2Cure works by directly involving crowds of people just like you.
Mark2Cure is a web-based, concept recognition, relationship extraction and validation tool for annotating text. It is intended to be a fun and exploratory way to help biomedical research, but can readily be extended to other disciplines of study where publication rates far exceed a researcher's ability to consume and incorporate knowledge from the literature.
Mark2Cure consists of 2 types of tasks:
- Named Entity Recognition (NER)
It is important to extract core concepts from an journal article in order for a computer to understand what the research is about. We're interested in having annotated text for Diseases, Genes and Treatments as we believe these three concepts are critical for understanding what the problem is (Disease), what may be causing it (Gene), and how might it be resolved (Treatment).
- Relationship Extraction (RE)
Understanding how Disease, Gene, and Treatment concepts are related is critical for determining what the research addressed. This provides data on how the same concepts are related across different papers.
If you are interested in volunteering as a Mark2Curator to read documents and help expedite our annotation process, please go to [Mark2Cure.org]. The information here is primarily for those with computer science, bioinformatics, and/or programming backgrounds who wish to assist with, develop, or build upon the code. If you stumbled upon this accidentally, we sincerely thank you for your interest in Mark2Cure and ask you to join us at [Mark2Cure.org] instead.
Mark2Cure is open-source and welcomes community contributions! Our issue tracker will soon be updated with issues which are readily "up for grabs" and will be marked as quick fix and/or help wanted.
To work on an issue in our issue tracker, please comment on the issue and provide the following information in your comment:
- Your intention to address the issue
- A rough time line from completion
That's it! We will assign it to you if it makes sense.
While we will always try to be responsive, our team is very small. We ask for your patience if we do not accept a pull request immediately. Our workloads fluctuate, which translates to inconsistent response times on these matters.
Please know that we are very thankful for your help and desire to contribute to this project.
Mark2Cure is distributed under the MIT License