October 15, 2009 (Vol. 29, No. 18)
Winners Receive $1,500 Prize and Invitrogen’s Neon Transfection System
Two scientists from Canada have figured out the answer to the Genetic Engineering & Biotechnology News Cryptogram Challenge. Adrienne Halupa, Ph.D., serves as project manager/research associate in the department of chemistry at the University of Toronto. Stephen Ross works as senior software developer at Toronto-based Devlin eBusiness Architects.
GEN, along with partner, Scintellix, and sponsor, Invitrogen, part of Life Technologies, congratulate Dr. Halupa and Ross on cracking the code embedded in the 96-well plate image.
In addition to being a co-winner of the $1,500 award, Dr. Halupa chose a second prize that was offered: one of three bench-top devices from Invitrogen. She selected the Neon™ transfection system for transfecting DNA and siRNA into any animal cell type.
“I think it’s the best option for our type of work,” said Dr. Halupa.
The answer to the Challenge was a statement by the famous Irish playwright George Bernard Shaw that “No question is so difficult to answer as that to which the answer is obvious.”
Peter C. Johnson, M.D., artist, and president and CEO of Scintellix, created the Cryptogram Challenge and said he “chose this statement for its ironic value in the context of a cryptogram.”
Stephen Ross and Adrienne Halupa, Ph.D.
Embedded Cipher
For the Challenge, Dr. Johnson embedded a cipher (algorithm for performing encryption and decryption) based on the cells in a standard ELISA plate.
“In this first edition of the Cryptogram Challenge, a follow on to last year’s ‘MicroArray Challenge’, I wanted to, once again, highlight the magnitude of information that is buried in biological imagery,” explained Dr. Johnson. “As experimentalists, we always try to reduce test variable and potential interpretations to the simplest form possible. The ELISA Cryptogram Challenge reveals that even when we do so, the potential for multiple interpretations remains.
“In the ELISA Cryptogram Challenge, I used a singlet code (each letter coded for by a single type of cell) but enhanced the difficulty by coding using a combination of specific RGB (red, blue green) intensities and also by requiring that letters and punctuation marks be designated within specific degrees of transparency. Once the transparency grouping was understood, the length of the statement lent itself to a solution using a standard frequency analysis. This is why the clues were given in their specific order. I congratulate the winners on their rapid solution.”
The Cryptogram Challenge is the first in a series of puzzles that will appear in GEN that will leverage the information represented in multiple types of biological experimentation readouts.
“Subsequent challenges will grow in difficulty and will be used to continue to highlight the enormous amount of information that we need to filter from our experiments in order to derive correct conclusions,” continued Dr. Johnson. “In this example of a multicolored ELISA plate, the visual beauty of our experiments can also be seen.”
How They Did It
To tackle the Cryptogram Challenge, Dr. Halupa watched the clues as they were posted on the GEN website.
“When the last one (clue #6) came up, allowing us to download the PowerPoint file, I immediately felt that we could solve the puzzle. I copied the numbers into a spreadsheet (grid format) and Stephen and I got to work on the puzzle.”
She first started looking for patterns in the numbers and made note of the frequencies of specific RGB codes.
“I also listed the transparencies associated with each instance of each RGB code,” she said. “I counted the number of times that specific transparencies occurred, essentially plotting a histogram. I noticed that most of the numbers were multiples of 10, but that a few were not. These codes seemed significant.”
To Ross, every one of the first five clues was essential to both indicate “what we were working with and to constrain the number of possibilities.” But, noted Ross, there were also key clues in the data itself: the transparencies, when set up as a histogram, looked like experimental data, which implied that values at the tails might indicate null results.
“And every nonfactor of 10 in the RGB codes corresponded with boundary values of the transparencies,” he pointed out. “Basically I tried a whole bunch of different ways of combining intensities and RGB codes using Perl scripts, until I finally realized by looking at Adrienne’s histogram of transparencies that I was going in the wrong direction and that the boundary values of transparencies should be treated as spaces. Then, with a bit more scripting, a traditional cryptogram emerged that could be solved in a straight-forward manner.”
Dr. Halupa called recognition of boundary values at the tails being null values a big leap in finding the solution to the puzzle. “A lot seemed to fall into place after Stephen did that,” she said. “The Perl scripts were also essential for the whole process to move quickly, since it allowed Stephen to crunch through a lot of possibilities.”
♦♦♦
CRYPTOGRAM CHALLENGE: ELISA REDUX
GEN has another contest called Cryptogram Challenge: ELISA Redux. The games began on Oct. 15, so register now if you think you have what it takes to be the next big winner. Besides a feature in GEN, you will also receive $1,500 and the chance to choose from one of five Invitrogen benchtop devices that range in value from $900 to $5,000.
