May 29Today was supposed to be the first day of work this summer. Yesterday was hectic as I was busy moving from Fort Collins to Boston. My parents came to drop me off, and we were spending the week with some extended family in Lexington (a suburb of Boston). I was exhausted, but then I received an email from Lenore telling me to come on Tuesday as the department would most likely be closed on Memorial Day. Thankfully, this gave me one day of reprieve. Instead of doing anything fun and exciting, I decided to tag along with my family to local grocery stores and do some Memorial-Day-BBQ shopping. We were out most morning, but came back at four to start grilling. As it was drizzling outside, I decided to go inside and read a book while waiting for the food to magically appear -- except I was so tired that I fell asleep. After dinner, I began to prepare (both mentally and physically) for the following day, and eventually drifted off into a restless slumber.
May 30Today was my first official day of work. I was both excited and nervous but mostly the latter. I was excited for this new prospect of working with Lenore and doing research in a completely new field of computational biology. I was happy to think that within a few hours, my mind would no longer be idle. Yet, there was an element of nervousness to my emotions as well. This was primarily due to the lack of confidence I had in myself. I did not know anything about computational biology and did not trust my rusty programming skills. I was scared that the other students would be more knowledgeable and intimidating. I was afraid of dragging the team down and not being a valuable asset to Lenore.
With these thoughts running to my head, I walked to Lenore’s office. She greeted me and my parents and gave us a brief tour of the computer science department. We chatted for a bit, then my parents said their goodbyes, and I got to work. She gave me a presentation and briefly talked about her research. She spoke about the diffusion state distance algorithm (to be written later).
After her talk, we went to Danish Pasty House. Along the way, I quickly made friends with the other students: Faith, Lily, Sam, and Rebekah. The food was great, but the best part of the meal was Lenore’s insistence that we have dessert there afterwards as well. It was only the first day, but I had already found a non-academic connection with my mentor.
After lunch, we headed back. Lenore wanted Sam and Lily to come up with a way to basically simplify the matrix that was the result of her DSD model, so that only three columns were left. I tagged along. Sam tried doing it using some online software (I think?) and Lily decided that writing some code would be better. However, we came across one problem. The input file, which Lily was trying to download onto her laptop, was too big. They then decided to do the task as well as some other ones on one of the department’s machines. I was confused as to what was going on exactly, so I decided to make myself more useful and started reading papers including one of Lenore’s most recent, yet to be published ones to get a better idea of what her research was in; and hopefully, what would be needed of me. May 31
I spent today morning doing a couple of things. I began by starting my website for this program. As I had never done so and did not know HTML, I followed the step-by-step instructions given on a site linked to the DREU website. Slowly, I was able to create files on a text editor and open them in my browser. So off I went: Lesson 1 check, Lesson 2 check, …, Lesson 8 no check. This was where it was shown how to link one page to another. I wrote the code they taught, but when I opened my “main” page on the browser and clicked the link, there was an error. I tried changing my code in various ways. One such way was by creating a folder and putting both the main page and the link to the following (let’s call it) subpage in it, and then explicitly typing the name of this folder and the corresponding subpage filename into the code for the link on the main page. On hindsight, I wonder why I did that – it wouldn’t have made a difference as they were both in the same folder before (as I only have one folder on my machine – the standard Documents one). I looked at other websites and changed a few more things, but nothing I seemed to be doing worked.
With that project on hold, Lenore and I did a few housekeeping stuff. Well actually one. We went up to the system administrator’s office and got my Tufts account set up, so that I could access the department’s server. After doing so, we went back to her office. She asked me how I was doing on my DREU program requirements, so I told her that I needed to get a website up and running. She responded with, “Oh that’s easy here let me show you.” She pulled up Terminal on her Linux machine, and began typing some stuff. I told her that my main concern was that I wanted to be able to do it on my laptop, so I could do it remotely. She then something along the lines of, “Yeah you just SSH it.” I had no idea what that was, so translated a few basic commands for me and told me to type along on my machine. In the middle of this lesson, I had a quick thought, "Could it be that Lesson 8 wasn't working because I hadn't yet published the website?" We created a file (index.html) and then tried to access the basic site http://www.cs.tufts.edu/~indrani. However, we were coming up with an error as the permissions were incorrect. We changed that, but even after doing so, it did not work. This time, however, it was because the department’s site was down as the department was doing some housekeeping.
That afternoon, Lenore told me that she had come up with a mini project for me. Knowing that I loved math, she said that I could split my day into two parts. The morning would be math and the afternoon would be programming. Of course I frowned at the latter part, so I jokingly said, “Well then, looks like I’ll be coming to and leaving work earlier.” She then gave me two graph theory textbooks and assigned some problems as well as a programming assignment from one of her classes. She gave me the option of either using a familiar language such as Java (which I learned three years ago in high school) or Python which she suggested was better. I immediately jumped to Java – there would be no way I was going to learn a whole new language. After listening to her project, I decided to get down to work.
The first thing I did was manually write down the idea about what my program will do to get an organized outline in my head. Here is a typed copy of my thoughts: - Have two text files. One is a list of protein name pairs and probabilities that they are connected. The other one has a list of proteins and functionalities which are just 2-digit numbers written next to them.
- The second list will be randomly split into two groups. The first group will have known functionalities and we will assume we do not know the functionalities of the second group.
- For each unknown node, we will check the functionalities of the neighboring nodes and vote for an appropriate functionality value. I will do so in two ways.
- If there is a confidence associated with a pair of nodes (from the first page), assume they are connected. Otherwise, don't. Assign the score of the unkonwn to be the same as the most popular functionality neighboring it. If this is not clear, randomly select a score from popular choices from adjacent nodes.
- Use the confidences in the first list to create expected functional values and use the most popular expected functional value (or randomly select one if it is not clear or there is a tie).
- Once estimates for the unknown functional values are found, they will be compared to the actual values (in the original list). Each time the estimate matches to one of the actual values, a counter will go up.
- The total (from the counter) will be tallied to give an accuracy percentage.
- The process will be done 10 times, and the accuracies will be averaged.
This bulleted list is incomprehensible, so I will polish it later on once I am done with the project.
This morning was hectic – I spent it moving into a small apartment that I would be sharing with four other people in an off-campus house. My parents and I pulled up to a house on Boston Avenue and walked up to the mailbox on the front porch and pulled out a pair of keys. I climbed up the creaky stairs and opened the door to the apartment. It was completely dark, empty, and smelled fishy (literally). All the doors were closed except for one, so I walked in and just assumed it was mine (it was). The room was spacious, but the furniture was old, dirty, and falling apart. I didn’t want to be late to work, so I decided just to unpack one suitcase which had some essentials. Thankfully, my parents were with me. While I was unpacking, they helped me clean up a bit. The lights in the room were also not working, so I would have to get some lightbulbs that would. The living conditions were not ideal, but I did not care as much as I generally would have, because I was happy for this was the first time in my life I chose and paid for a place to live by myself. I was slowly becoming independent from my parents. After this, I took my parents to Danish Pastry House, and then I went back to Halligan Hall (the department).
Afterwards, all of us working with Lenore touched base and had a meeting with another group led by Lenore’s collaborator, Dr. Donna Slonim. We all went over introductions and briefly talked about what we were working on. Once this was done, Lenore stood up and gave us a lecture about their recent and future work on clustering proteins based on their closeness which is found by the DSD algorithm as well as using DSD to categorize the proteins based on closeness and then clustering them. She also talked about what she and the Tufts team did for the DREAM competition. I found her talk to be exciting, primarily because I kept thinking of new questions were currently being / had not yet been researched.
After the talk, it was pretty late in the afternoon, so I decided to relax a bit and do the math problems on graph isomorphisms she had assigned the day before.
By Friday, I was already exhausted from the previous week and ready for the weekend. I came into work a bit down-in-the-dumps as my parents were leaving. I did basically three things all day. Actually, come to think of it, I spent most of the day recapping topics I learned in high school that I thought could be applied to work here. I started off doing more of Lenore’s math problems. Some of the problems were straight-forward exercises, but for a few, I tried to find out more complicated but elegant solutions. For instance, one exercise was to show that there are 11 non-isomorphic simple graphs on four vertices. I, for some reason, thought that this could be shown by showing permutations on groups, so I went back and looked up some things I learned from abstract algebra three years ago. I got too deep into memory lane, so I decided to stop the tangent tracks and just come up with a simpler solution. As for programming, I hadn’t yet installed Java and NetBeans (the IDE I was used to using) on my computer, so I decided to outline the various components (classes) of my program in my notebook. While doing so, I had to look up different things to refresh my memory. For example, I had to rehash how to read inputs from files and use the Scanner package as well as other relevant commands. The third thing I did was begin to write my log of what I did this past week for this DREU website.
Lenore had forwarded us four undergraduate students a flyer about a
free RiverFest happening in Boston today, so the four of us decided to go.
Sadly, however, Sam was sick, so the three of us decided to go. Before going,
Faith and I decided to eat at Masala (an Indian buffet) and meet up with Lily
(who had already eaten). However, we were taking so long to eat (what can I say
they had a huge selection), that Lily got tired of waiting decided to stay. As
it was raining, the two of us decided that we would rather stay and hang out in
Medford. Faith showed me around Davis Square and we walked into a few cute
stores. On an impulse, we decided that we wanted to watch “Wonder Woman”.
However, we missed the start of the showing by 10 minutes. We then decided to
go back to my apartment and chill out for a little bit before coming back for
the 4:45 showing. We hung out for a bit, before heading back to Davis Square
again. Sadly, on our way, we realized that we would be late. Then, we targeted another
showing. This time, we made it there on time, only to realize that they had
sold out. Oh well, we would just have to leave “Wonder Woman” for another day. Anyways,
I still had fun.
My plan for today was to finish unpacking, go grocery shopping, and
catch up on work. Nothing exciting happened. Grocery shopping was hard in three
ways: I did not know how to cook, so I did not know what to buy; I had to walk fifteen
minutes carrying two unwieldy brown paper bags filled with groceries that did
not have handles; and I only had half of a shelf in a tiny fridge to put my
groceries, so I could not buy things in bulk. As a result, I probably did not
buy enough food, and will have to go shopping again in two days.
|