 |
The Laboratory Robotics Interest Group Home |
|
|
|
[web_discuss/_private/disc3_ahdr.htm]
A 384-well mathmatical puzzle.From: Kevin Olsen, Wyeth Ayerst Research CommentsDear Fellow LRIG Members, Like many of you we are making 384 well microplates on a routine basis. Most of the 384 well plates have been made by combining aliquots from four 96 well plates. Each 96 well plate has a five digit number (if made internally) or a six digit number (if made by an outside collaborator.) For example, suppose we combine aliquots from the following 96 well plates: 12345 12346 12347 12348 The problem is, what do we call the new plate? If we call it 12345-12346-12347-12348, the label will be longer than the plate itself. If we give it a new number entirely, for example 54321, then there is no easy way to reference the original plates. So here is the puzzle. Is there a quick algorithm whereby the original four numbers can be combined into one new number? And once this is done, can this new number be readily converted back to the original four numbers? (Using two-dimensional bar code labels is not allowed.) The same problem occurs when we make 1536 well plates by combining aliquots from 16, 96 well plates. One possible solution was proposed by Bart Zoltan of our Biomedical Engineering department. The new number’s first five digits are the ID of the plate with the lowest ID number. The next three digits are the increment from the last plate. Thus in the example cited above, the number of the 384 well plate becomes: 12334-111 It's a really elegant solution because the operator can mentally convert back into the original plate numbers very easily. But the problem arises with a set of source plates with more than 9 numbers between any two of them. What happens if a screening laboratory decides that they want to pick any four numbers out of several hundred? If the subscribers on this list are unaware of a suitable algorithm, does anyone know any cryptographers? Thank you, Kevin Olsen Wyeth Ayerst Research Compound Bank Pearl River, NY, 10965 914-732-3392 [web_discuss/_private/disc3_aftr.htm] |
|
|
with questions or comments about this web site.