How to increase the overall success rate of experiments - (Mar/01/2013 )

How can you increase the overall success rate of your experiments ?

A strange question perhaps, and I know it's normal a lot of experiments just don't work, but as mine seems to be lower than usual (there's always something ), I'd like to hear your ideas and thoughts.

I often feel the same way -- especially the "it's always something" part! My answer to your question is to maintain the highest level of consistency across all facets of your experiments - reagents, cells, protocols, timing, etc. In other words, do everything the same way, every time. And record it, of course. It's a simple concept but not always easy to put into practice. I haven't quite mastered it yet myself.

Reading papers (to get an idea about the background and from this an idea about the mechanisms and possible error sources) and protocols before, asking or watching colleagues/bioforum, exact planning.
Training.
Consequent thinking what has gone wrong if the experiment failed in order to tweak the causal settings and not just arbitrarily trying out new approaches.

Think about whatever you are doing. Even if it seems mundane and repetitive, do not just add solutions blindly. Whenever you make a move think about it 3 times. It may sound tedious, but it will save you time in the long run.

It is also important to understand what each step entails and what ever chemical does.

it always happened to me i cant get a quality result at the very first time i plan and do the experiment.

what i learnt is, follow the principle rather than blindly following the protocol...unless an experiment got used in my hands, i always use to have a positive control along, to check whether the technique by itself worked or not..

dnt repeat the experiment unless you spot the possible error in the previous experiment, i believe the failed experiment definitely needed a change in its protocol or the way of doing.

Then organization, which i think is one of the major factor behind the success rate, i always try to find where all i could put a break or possible to extend overnight in any technique, so that i wnt rush any steps to finish on the same day...

For me, the pilot/preliminary experiments are very very important. This will provide you the foundation to your big experiments and give you an idea of how your experiment will unfold. And I agree with other posters in this thread - understand principle of what you're doing instead of following protocol blindly.

I would record down my steps (especially in tedious long protocols such as western blot) so I can trace back to identify where it went wrong or needs optimisation.

I also find that the smallest things make a big difference sometimes, e.g. buffer pH, salt concentrations, supplement calculations, temperature, and cell culture technique. These might or might not have direct relationship with your experiment but could influence the final outcome.

It's also important to plan the experiment with your boss or seniors before you get into it. The last thing you want is to start a badly planned experiment that leads to nowhere.

It depends on why your success rate is what it is. i suppose there are two reasons why an experiment goes sour. (i) error in execution and (ii) error in design.any experiments require careful planning. having stated the obvious, i do all my thinking/planning before the experiments, write down exact game of plan before you actually execute. once i am doing the experiments, i just follow my protocol exactly. i always tell my students to think about the experiments. what i mean is that i picture myself doing each steps of the protocol the night before. however, i have to admit that sometimes i found myself doing something terribly wrong. then i guess you have to think on your feet to figure out the problem. to develop skills to think on your feet, it is important to know why you are doing each steps. when you know why you are doing what you are doing, it is easier to fix your mistakes. if it is an experimental flaw, then you are doomed to begin with. it doesnt matter how well you have executed your experiments, if you are missing controls, or going after leads that are false, what is the point?

There is a learning curve to this, and it is a steep one. It's sometimes too pricy to learn by trial and error, but your perseverance will one day translate to good science.

i dont usually believe in god, but if you are up there, please save me superman! --homer simpson

my best,

jeff kwak

Thanks for sharing your ideas....

Some simple implementation things:
* Keep a clean bench. No old stuff. When you finish with a reagent or tool, put it away immediately.
* Use physical location of samples to track the progress of adding reagents, e.g. -- move the tube down or left when you are done with it for this step.
* Use pipet tips from the box in order and related to what you are doing, so you can know whether you are on this sample or the next.
* Count. Always count. Know where you are.
* Do not let others interrupt you. This almost invariably causes errors.
* Use master mixes to reduce the amount of pipetting and increase consistency
* Think about what you are doing at each step, and why you are doing it. This only works if you really understand the protocol.