I spent the first half of the week trying to prove my claim about complete graphs, but I did not really get anywhere. Lenore suggested that I explain my problem to other students to get different perspectives and clarify my thought process. We planned to do this next week. I then focused my efforts on the programming project. I realized that I had been over complicating things by storing all of my data (genes and functionalities) as arrays of arrays with string variables. This was becoming hard to work with and manipulate, so I decided to start over and use a different object type (in this case dictionaries). I started my program from scratch and more or less finished it by the end of the week. However, I had one problem -- the voting algorithm for which an unknown node's functionality was predicted by its neighbors functionalities was not strong. I was simply looking at the neighbors and casting a majority vote (and breaking ties by randomly selecting one). I did not take into account the confidence scores (relating to the probability that two nodes were connected) in my protocol.