Weekly Journals

Week 1 ( June 1st ~ June 7th):

 

    This week, I have been reading the AI book written by Russell -- coincidentally, a professor from Berkeley, my home school I went through various searching algorithms, such as A* search, Memory-bounded heuristic search, Greedy best-first search...etc. I am informed that I will be working on the trajectories of protein folding, but I need to wait for another intern to come to have a meeting where I will be able to learn more details about the project I am going to work on. Meanwhile, besides the AI book, I have also been reading the Introduction to Protein Structure by Branden and Tooze. Both readings tuned out to be pretty interesting and delightful. The first one introduces lots of elegant searching algorithms in robotics, and the second provides insights into molecular biology, especially the mechanism of protein folding.

 

 

Week 2 ( June 8th ~ June 14th):

 

    We had our first meeting this week, and Amarda asked us to learn about achievements of other researchers so far in trajectory analysis and critical even detection. Both of us soon learned that the task is not as easy as it sounds. I searched the electronic database and found papers and books on trajectory analysis, but only a few are actually related to the project that I am gonna work on, and furthermore, among these few articles, the algorithms seem pretty elementary, more or less simliar to the trajectory analysis methods that have already been used in mechanical physics lab. On the plus side, the limited resourses we found seem to indicate that there is a lot of potential of exciting work in this field.

 

 

Week 3 ( June 15th ~ June 21st):

 

    After the weekly meeting, Amarda decided that we are going to spend this week in learning about geometric modeling, while she starts the frame work for the project. Due to the complexity in the protein structure, many concepts that I have came across in linear algebra, such as basis, span, and projections, turn out to be closely related to GM, and have become handy for me. This week, I also learned about the PCA (Principle Component Analysis) technique in data analysis from the statistical perspective. It seems that this project, which is very interdisciplinary in nature, will be very interesting. I am looking forward to it.

 

Week 4 ( June 22nd ~ June 28th):

 

    This week, I went to the intensive two-week summer course, Biology for Modelers, at Rice. Lecture is from Monday through Friday, plus lab sections on Tuesday and Thursday. The course turned out to be very helpful. It covers a variety of subject from enzyme biochemistry to Prokaryotic genetics, deepened my understand of the detailed structure and function of proteins and much more beyond.  This week, when alone, I also taught myself the HMM (Hidden Markov Model).  Learning HMM by itself is fascinating to me, because the model is so well structuralized and amazingly powerful. But on the other hand, I felt the uncertainty that one can feel when walking in the dark -- HMM is widely used in speech recognition, but whether or not it will be useful in our project, at this stage, no one is able to tell. It reminded to me what Albert Einstein said, "If we knew what it was we were doing, it would not be called research, would it?" The drawback of the HMM is that it requires all inputs to be specific, which is not always possible when studying protein folding. I will read more and see if it will eventually be a possible path for our project. 

 

                                                                                                                                                                                                                                                         Next Page