I was going through some files and came across this image I made in July, 2006, using only Photoshop and the trackpad on my Macbook (I hadn't started using Illustrator yet). My postdoc advisor mentioned that it would be nice to have a figure depicting the system we were developing - heterobifunctional ligands that would bind to both a decavalent IgM antibody and the carbohydrate-binding cell surface receptor, CD22. Because CD22 engages its ligands with such low affinity, multivalency is needed to acheive stable binding. Subsequent to this figure, I made many cleaner variations using vector graphics in Illustrator, but my advisor refused to replace this one in his talks. To me, it represents the beginning of my serious consideration of actually doing science illustration for a living. Not only did I have a blast drawing it, I experienced first-hand the utility of trying to draw what you think is going on in the eppendorf tube. It forced me to really think about the relative concentrations of everything and how the receptors might cluster. It led me to search the literature until I found out that the Fab fragments bend down as though on a hinge to engage their cognate antigens, as shown. Prior to that, we had been drawing the bound IgM as more of a tee-pee shape, or like the claw from that bowling alley game that theoretically retrieves stuffed teddy bears with the claw but is more likely to just trap a small child within its walls. Likewise with the projects I take on now, what I'd much rather hear than, "Ooh, it's lovely!" is, "By jove, I've never thought of it that way!".