Research collaborations have a synergistic effect that often yield good results, so they are always encouraged.However, creating and organizing a strong research group is a difficult task.One of most main concern of individual researcher is how to find collaborators that match his expertise best.In this paper, we proposed a method to make link predictions in co-authorship network, where nodes represent authors and links denote collaborations.Topological features in network such as Adamic/Adar, Common Neighbors, Jaccards Coefficient, Preferential Attachment, Katzβ, PropFlow, Shortest Path count, and RootedPageRank might give good suggestions for future collaborations.First, topological features were systematically extracted from the network.Then, supervised models were used to learn, the best weights associated with different topological features in deciding the co-author relationships.We tested our model on a co-authorship network in Coronary Heart Disease research domain and obtained encouraging accuracy.This made us believe that our models could be helpful in building and managing strong research groups.