• Question: how many of your experiments have gone wrong in he past? and how?

    Asked by to Claire, Ian, Sergey, Vicky, Zena on 17 Jun 2014. This question was also asked by , .
    • Photo: Claire Shooter

      Claire Shooter answered on 17 Jun 2014:

      I had a terrible problem last year where I was doing quite a complicated DNA preparation experiment. At the end of my experiment, the DNA would disappear. I kept thinking that I must have done something wrong and not realised. Eventually I realised that the room I was working in (which had no air conditioning) was actually too hot for my experiment to work properly!

      There are dozens of different times I’ve messed up experiments at work. It’s part of the learning process and it isn’t the end of the world. The most important things in the lab are to plan ahead, think rationally and persevere – not just to get everything right first time.

    • Photo: Sergey Lamzin

      Sergey Lamzin answered on 17 Jun 2014:

      Pretty much all of them. My experiments involve making software and I have to suffer the same way as any other software developer:

      – Spend an hour writing code
      – Find out it does not compile [failed experiment]
      – Spend an hour fixing syntax mistakes
      – Find out it compiles, but parts are missing, so it doesn’t run [failed experiment]
      – Repeat the above two to tree times until I get a program
      – Run it until it crashes [failed experiment]
      – Attach a debugger and find where it crashes
      – Repeat the above 20 times until it runs and does not crash
      – Run it, but get nonsense results [failed experiment]
      – Spend a week trying to figure out why it is not producing proper results, repeating the above until it does
      – Finally it runs and does what it is supposed to [successful experiment]

      Reward: 3000 XP, Back pain, square eyes, more work!

    • Photo: Ian Simpson

      Ian Simpson answered on 17 Jun 2014:

      Lots. Experimentation is exactly that, although there are recipes to follow a lot of it comes down to experience and trial and error. I’ve not had any massive problems, most things that go wrong are quite trivial yet still manage to cause havoc.

      I remember my first day in a wet lab very well when I was an undergraduate. I was pouring an agarose gel (which is basically very high quality jelly) that we use to separate DNA fragments under an electrical current. I wasn’t happy with the gel so I poured it when it was still molten down the lab drain along with a lot of cold water to dilute it and carry it away (this is harmless by the way, although nowadays we would let it set and put it in a special waste disposal system). Little did I know that my new lab was having problems with the internal drain pipes so the flow was low. The gel set solid in the pipe, the water built up and pulled the entire drainage system off the wall, crashing onto the floor of the lab. A bit unlucky but these things can happen !

      My favourite thing that went wrong was with one of my early PhD. students who had spent the best part of the week preparing a fairly complicated blot to measure the levels of some proteins in a cell extract. At the end of the procedure you expose the gel to some X-ray film at -80C for a few days and then finally develop the picture from the film (like old photos !). After all this work she accidentally binned the film and put the blot through the automatic developing machine instead. It looked beautiful though I have no idea what it showed.

      As you get more experienced you just get more careful and methodical and you tend not to make so many mistakes. Some people definitely seem to have a talent for the dramatic in the lab, normally involving pipettes, scalpels and naked flames, but most are safe and successful.

    • Photo: Vicky Schneider

      Vicky Schneider answered on 18 Jun 2014:

      good question…lots cause you often find unpredictable factors that interfered and that you only can see by trying…so its all about trial and error and improving experimental design as you go 🙂

Comments