This post was originally targeted toward undergrad/graduate students interested in building a LIMS to help conduct certain experiments but it can be useful for business analysts and others entering the field.
Today we are going to discuss, at a very abstract level, what a LIMS is and what it does. Before going too much further, read the following rubber elasticity and temperature experiment (don't worry, it was written at a high-school level).
More than Just Data
The belief that LIMS are mainly about data aggregation and storage is largely a misconception. There are many different aspects that simultaneously do the work of the system. In the rubber band temperature experiment there is, of course, data -- three trials consisting of temperature readings spanning from 3 to 45 degrees Celsius in 3 degree increments and several 'constants' that are really measurements relating to the materials being utilized in the test itself. But there is much more!
What Do You Get
The LIMS does many things for you, many that you do not immediately notice. Here is just a subset:
- 1) Access control - since the LIMS only allows approved users to get into the system there is a separation between 'laboratory personnel' and everybody else.
- 2) Equipment/Materials Identification - A rubber band was used in the experiment -- but which one was it? In a LIMS we can uniquely identify the materials used in the experiment. This means that we know whether Trial 2 and Trial 3 were done using the same rubber band/experimental setup rig or whether they were all done separately with different rubber bands.
- 3) Equipment Management - Equipment breaks down or gets damaged with daily use. The LIMS will often have ways to identify good/bad equipment and allow for it to be placed out of commission in order to avoid using it in subsequent experiments and ruining results.
- 4) Work Assignments - Although all three runs need to be done with the same setup and rubber band there is no requirement that they be done by the same person. The experiment can be run by different people on different shifts by publishing some form of worksheet or list that assigns the work.
- 5) Status/Timestamps - Some of the most popular aspects of a LIMS are the status values and timestamp captures that occur largely 'sight unseen' by the users. In a typical LIMS every piece of information actually has this metadata associated with it. Consider the first temperature reading in the first trial (Trial 1) of 16cm. In a LIMS when that information gets entered the system might flag that sample associated with it as 'In-Progress' and capture the exact time the value was put into the system. Later on, at a glance, one can identify whether all of the trials were run or if there were missing data points.
- 6) Pre-Defined Data Points - In a LIMS the only data points that could be captured would be those specified in the data section. If the lab analyst tried to enter 4 degrees Celsius, for instance, the system could reject the attempt.
- 7) Authorisation - At the end of the experiment someone must review the results and determine if everything went according to plan. Let's say that Experimenter A missed 1/2 of the data points on Trial 1 or the results appear entirely random. Another trial can be scheduled and recorded in this case and the errant trial cancelled, making its results void in the system. This review itself is captured in the system as well as the determination that the results captured were or were not considered acceptable.
- 8) Hazards - Although this experiment seems fairly benign there is a slight possibility of rubber band breakage. Perhaps eye protection should be worn. A LIMS would tell you what type of hazards are present when dealing with materials.
As you can see, lots of the features of a LIMS have to do with planned collaboration. If a single researcher were acting alone there would be less need for timestamp capture, pre-defined data points, authorization, etc. That individual would simply 'remember' what to do and whether the results entered should be considered correct. There would be no need for a work assignment because all the work would be expected to be performed by a single individual.
Collaboration is slower than doing everything yourself, but that is the price you must eventually pay for repeatability and verification of results. One could argue that the very purpose of a LIMS is to reduce variation needed to successfully repeat experimental processes with a great degree of success (there's a definition you can use). This helps both the reviewer and the analysts. If there was a great degree of variation between the analysts the reviewer's job would be nigh impossible. Reducing variation makes the reviewer's job easier. It is the reason judges want to hear from barristers that speak the legal 'language' rather than the layperson; the reason why we elect representatives rather than all try and squeeze into the town square for a vote. In a LIMS there is a tacit understanding that everything is eventually going to have to flow upstream to a reviewer for a 'yea/nay' decision. Reducing variation is a strategy for making such decisions more quickly. In the case of rejections some steps are going to have to be repeated or some constraints in the system will have to be modified. These things are not unique to LIMS, they are a part of everyday life!Go Back
Citation: LIMS Overview. (2015). Retrieved Wed Mar 22 22:06:32 2017, from http://www.limsexpert.com/cgi-bin/bixchange/bixchange.cgi?pom=limsexpert3;iid=readMore;go=1439734552