I made the saturation game to teach first year undergraduate students how small molecules, like drugs, bind to proteins. Some students understand this through mathematical models, but others need it explained in a more tangible way. My students are diverse; there is a wide range of incoming qualifications and preferred learning styles, and over 50% are non-white. By starting a teaching laboratory with a card game I helped different kinds of learners draw the correct conclusions from their experimental results.
In the practical the students measure the binding affinity of a protein (avidin) for a small molecule (HABA) by taking a sample of avidin and repeatedly adding small amounts of HABA. After each addition they measure the proportion of avidin molecules that have bound to HABA, recording a % saturation value. Figure 1 represents the point at which avidin is 50% saturated.
Figure 1. Half of the avidin molecules (blue blobs, with concave binding sites) are bound to HABA molecules (orange triangles).
Figure 2 shows that at the start of the experiment, adding a little HABA leads to a big increase in % saturation. At the end, though, adding the same amount of HABA leads to a smaller increase, as the saturation curve levels off. The saturation game teaches students why this happens, correcting a common misconception.
Figure 2. Increasing the drug concentration leads to a hyperbolic increase in protein binding: the saturation curve.
The saturation game
I used a TurningPoint MCQ to anonymously ask a total of 92 students, before they played the saturation game, why adding the same amount of HABA would lead to a smaller increase in % saturation at the end of the experiment. Only 29 of them (32%) chose the correct reason: that the added HABA is less likely to bind to avidin than it was at the start, because the previously added HABA has blocked some of the binding sites. 49 students (53%) wrongly believed that the added HABA would bind to the protein with a lower affinity than it had at the start.
The students then played the saturation game, which is a kind of patience. They played in pairs: one student dealt the cards while the other recorded their results in a Google Form. Each time they played a card from their shuffled, standard deck it represented the addition of HABA. In the first round they dealt the top card from the deck face-up onto the bench and noted that one card had “bound” to avidin. For each round that followed they examined the next card in the deck. If the value of the card was different from all of the cards currently “bound” then they laid the new card face-up onto the bench and recorded that one more card had “bound”. If it matched one of the cards on the bench, they discarded the new card and recorded an unchanging number of “bound” cards. Two groups of students played 25 rounds of this game, which took about 15 minutes.
When the students had finished playing the saturation game I showed them that their collective results formed a saturation curve. Figure 3 shows that when the saturation game is played many times it gives a “binding” curve of almost exactly the same shape as the ideal protein binding curve.
Figure 3. An ideal protein binding curve (orange) compared to the ideal saturation game curve (black). Probability calculations were performed by Dr Alastair Kay.
With no further comment I then asked the students to answer the same TurningPoint question that I had asked at the start. Only 18 students (20%) now wrongly believed that the protein’s binding affinity changed as the experiment progressed. Simply playing the saturation game had reduced the rate of the most common misconception from 53% to 20%.
50 students (54%) chose the right answer, reporting that the added HABA is less likely to bind to avidin at the end of the experiment. These students understood that shuffled cards were less likely to “bind” near the end of the game because previously dealt cards had occupied some of the “binding sites” on the bench. Furthermore, they had seen that the game produced a curve very similar to an ideal protein binding curve.
Now that most students understood the stochastic basis of the protein saturation curve, I briefly explained the saturation game and emphasized its correct interpretation. I then invited the students to begin their practical. The average grade for the post-lab assessment was 68%, suggesting a very good understanding of the experiment.
Academics often equate pedagogical gaming with gamification, but the two concepts are distinct. The saturation game is an example of a game designed to teach by analogy, like a card game “parable”. There are other topics, particularly those involving probability, that can be similarly illustrated with cards and/or dice.
A handful of students who were persuaded by the saturation game to relinquish their “lower affinity” misconception landed on a different misconception instead, as revealed by their answers to a distractor in the MCQ. I will try to stop this happening in my next iteration of the game.