Track your lab data from when it's just an idea.
*** A quick note: I’m going to start occasionally including Sponsored Highlights at the bottom of newsletter posts, where I’ll write about software that I think deserves more attention.
This week it’s for Kaleidoscope, a tool that I’ve written about in the past and continue to be excited about. Last week I hosted a webinar with Kaleidoscope’s Bogdan Knezevic, and you can find the recording here. Or to learn more about what they do, read the highlight below. ***
Last week, I wrote about one of the few places where the Biotech Reference Stack is intentionally opinionated, namely defining a separate staging area for lab data and metadata. This week, I want to talk about a second one: The third component from the top in the record column, “Placeholders for data from upcoming experiments.”
This component is at the very top of the process to emphasize that as soon you have the vaguest thought of a dataset you might want to generate to answer a particular question or make a decision, you should write it down in a central place. Give it a name. Make a note about what it’s for and what it should look like.
Most of the teams I talk to do this at much later stage, if at all. Early on, they have informal conversations about upcoming experiments. Sometimes they’ll sketch out what the data will look like, but more often they just allude to it. The experiment plan may be captured in an ELN or another system, and an expert could theoretically deduce from that what kind of data to expect. But there’s a big difference between information that a bench scientist could deduce intuitively and information that the entire team can readily find and read.
Informal conversations about things that don’t have formal names can work in small groups with tight communication that can keep informal names relatively consistent and figure things out from a shared context. That’s why bench teams have evolved to work this way - for the kinds of work that has traditionally driven biotechs, that’s the most efficient way.
But once you have to get more teams involved - computational biology, automation, machine learning - you quickly lose the tight communication that makes it work. With this more complicated network of teams and functions, biotech organizations need to be much more deliberate about communicating. And that means giving upcoming datasets (in addition to experiments, assays, etc.) formal names and storing them in a consistent, central place.
This central database of upcoming datasets allows these other teams to not only plan for upcoming work, but to contribute to the design of the experiments that produce them. One of the biggest complaints from computational biologists I talk to is being asked to perform miracles on data that was generated with too few samples, a too-shallow sequencing depth, the wrong cell line, etc. Catching that before the experiment starts can save weeks or months, not to mention the financial cost of a wasted experiment.
Sure, teams can find these kinds of issues through less formal meetings and discussions, and they can track work across sub-teams without a central place for this information. It’s just a lot more work, and less reliable.
I believe that as biotechs get used to the idea that they need to deliberately track this kind of information, and as they find better tools and processes to do it, this is going to be much more common. So I put it in the Reference Stack both to try and fit where things are going, and to maybe nudge it along a little.
Sponsored Highlight: Kaleidoscope
Drug Development involves juggling work between lots of different teams. As we explored in the webinar last week, small misalignments between these teams can lead to big delays that cascade into months added to your IND timeline and cut from your patent clock.
Kaleidoscope helps eliminate these delays by collecting the most important data from all your systems in one central place, for easy review, actioning of next steps, and team hand-offs.
R&D project planning tools (Gantt, Kanban, Tasks, and Experiments) show stakeholders what teams are working on, when they can expect results, and which timelines need extra attention.
Customizable data dashboards and cross-team experiment searching/filtering allows team members to quickly answer simple questions that previously took days to track down, letting them get back to the science.
Centralized data review and experiment planning views allow teams to quickly determine the most impactful next step and record it at the appropriate level of detail.
Kaleidoscope ensures that everyone in your biotech can consistently work on the most important problems and data at the earliest possible moment. To learn more, check out the webinar or head over to kaleidoscope.bio.