Hacker News new | past | comments | ask | show | jobs | submit login

With an n=83 they did clustering, random forests and a regression model to develop a new genetic test? At least they used the standard questionnaire which should resist the chance of them overfitting the data.

But, the problem could be generally solved by treating people with the disease just using the questionnaire. Taking people's blood to see if treatment is progressing is an interesting way to assess people for treatment. Would be great if this test was expanded to a larger amount of people.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: