Cost and benefit analysis.
Its very simple. In biochemical and biomedical research labs we have space, equipment (pH Meters, centrifuges, PCR machines, pipettors and more), shelf chemicals (dry and liquids), biochemicals (proteins, enzymes, metabolites, DNA, RNA), bacteria, mammalian cells, culture media, and people (students, staff, research assistants).
In the academic research environment, as a Primary Investigator (PI) and a Professor, we have to write grant applications to government agencies (NIH, CDC, NSF) and private foundations to get the cash that will buy everything listed above and also to pay our assistants and payed student research assistants, and sometimes to pay a large hunk of our own academic year salary. (Most professors at Universities get paid based on 8 months, Sept to May, which is the “academic salary”. Any pay for working during the summer usually has to be obtained from a research grant, and if there is not enough cash for that then we work during the summer too, for free.)
The ability to collect results, solve research questions, train students to be the next generation of scientists and physicians, is directly impacted by how far we can stretch our funds. In the end it all comes from the same pot of money we have managed to convince an agency to give to us. So its important, especially if you are a new assistant professor, to spend your money wisely. Wisely does not necessarily mean purchase the cheapest option on its face value. That can be a mistake that leaves your research dead in the water too.
My advice: spreadsheets. Use them. Everything you do in the lab involves five things usually:
- person power (hours and wages/hr)
- labwares (tubes, pipettes, tips, dishes etc)
Each of those five items in the list has a cost associated with it. So use a spread sheet to compare alternative methods and processes. Sometimes you will be able to consider doing a process in-house or out of house.
Need a new antibody that you want made to a specific amino acid sequence from a protein? You do that in the 2000’s by having a company make it for you. They make the epitope, inject the rabbits, collects the serum and purify the IgG antibody to your target of interest all for about $1200. If you do that yourself, you will spend several thousands of $$ in time and reagents and rabbits to make the same product. So, antibodies you make with a commercial provider. That is an obvious example of a cost benefit analysis in your research.
Sometimes the cost benefit is not that obvious.
Real time PCR is an example. My lab is expert at gene expression analysis, very accurately, using real time PCR. Investigators at my University and from other universities, have asked me to guide them as they started to measure expression of their gene of interest in their systems of study. These days I would recommend using Taqman probe chemistry and an endogenous control probe set. What that means is not important for this discussion, but suffice to say that it is true that Taqman chemistry appears to look a bit more costly than the method just based on Sybr-Green. Sybr-green just adds a dye to almost any PCR reaction, and technically one can get some cheaper PCR reagents to do this method. Seems cheaper, and many labs will go for the Sybr-green option regardless of my advice. I know that in the end the Taqman probe method saves money compared to Sybr-Green. Why? Simple. While each individual reaction is a bit more costly in Taqman chemistry versus Sybr-Green, Taqman chemistry allows running an endogenous normalizer control in the SAME reaction tube. Thus we can set up half as many tubes as using the Sybr-Green method. In the later, a separate reaction must be done for the gene of interest and the normalizer gene. Futhermore, there are data quality issues to consider. Sybr-green will pick up undesired PCR products, “false positive signals”, and the reaction conditions will then have to be adjusted and the experiment repeated entirely. Taqman chemistry does not suffer that problem, and the chemistry is also very specific to specific mRNAs expression from genes. Thus we rarely need to repeat an analysis for failure reasons. That saves money.
“Garbage in, garbage out” is important.
You can save lots of money by spending less of it on an experiment, but if you get garbage out for data, you just have garbage. No results. Eventually you can burn through your pile of research cash, processing a lot of garbage while thinking you are saving money. So for experiments, think about the quality of data you need in the end. Better quality data is better data to publish, period. Its easier for reviewers to evaluate for conclusions.
Another place to save money is software for data processing.
Most computational tools available start out free, from academic researchers and professors. Then companies make programs with the same programming methods, add more graphical user interface and sell it back to researchers for hundreds or thousands of dollars. Graphing and curve fitting software for one example. However, if you take a little time here and there and experiment, you can actually find many software solutions that are free and shared. For dose curve fitting, as an example, you can get a very powerful stats environment called “R”. There is a dose curve fitting package you get which runs in the R system, called “drc”.
The R-project can replace SPSS, very expensive statistical tools software, and is now used in all fields of research from sociology to genome-wide association studies to dose curve fitting. You have to install a few packages yourself (on a Mac, we use Macs), but the installer packages do the work, and you have to get used to typing some terminal command line instructions. But what you get out is advanced high quality data fitting. You have the benefit of saving funds, to pay more students, who need tuition money badly and everyone is a winner. Even your granting agency enjoys seeing you get more bang for their buck.
MyCurveFit.com is a nice tool, web interface with an excell plug in optional, for free occasional fitting of your dose response data. However, they limit how many fits and much data points you can fit in their free version. Taking off those limits requires a monthly subscription.
Better Curve Fitting , more options, free, but you need to type more:
If you are a biology lab and want to just fit dose response curves, email me, I can point you to the packages to download and install and how to fit your own data from an excel spread sheet. firstname.lastname@example.org