Passing DNA through a silica column twice does not increase yield – A cautionary tale of cognitive bias

Scientists are stereotyped as objective thinkers, putting hard facts above our own subjective experience. If you’re being honest though, as a group of people we’re particularly vulnerable to a host of cognitive biases including magical thinking, confirmation bias, etc. Why the maudlin tone??? Well, I challenged my own magical thinking, and it turns out I was wrong. Not just a little wrong, but pretty definitively wrong. Oh well, better to don the dunce cap for one day, than be a dunce forever.

We all have tips and tricks, passed on and accumulated over the years. One of mine was regarding the use of silica columns, the idea that passing your DNA/chaotrope over the silica spin colum TWICE would improve yield. I had heard about it from a protocol or forum, lost to time, but I have been doing this for years. Makes sense right, what if there’s leftover DNA in your flow through, passing it over the silica a second time might help you out, right? RIGHT???

So there I was, twenty minipreps on the go, passing the lysate through the column a second time. Plenty of time for the mind to wander…

Brain:
“…this sucks, twenty minipreps in one sitting sucks, passing the lysate through a second time sucks even harder…”
“…is this really necessary?”
“…you’ve been doing it this way for years, you don’t want to lose yield, do you? HUH? DO YOU? YOUR GRADUATION AND LIFE AND CAREER RESTS ON THIS!”
“…boy would I feel stupid if I’ve been wasting my time for years…”

I’ve been having that conversation in my head for a while now, each time shying away from testing my hypothesis, BELIEVING that it made sense, that it had to work. But really, I was just afraid of feeling stupid. As more time passed, the stupider I would feel If I was wrong, the more I put off testing this. What’s worse than deceiving yourself? Deceiving other people and passing down techniques which aren’t grounded in reality. So, it was time to challenge myself.

Experimental setup? 50 ug of maxiprepped plasmid DNA dissolved in 1000 uL TE. I added 5 volumes of DNA binding buffer to the DNA, then from the master stock I added 600 uL of DNA/chaotrope to 10 individual silica spin columns. Five columns I centrifuged the DNA through once, the other five I took the flow through and passed it through the column again. I eluted with 100 uL of warm elution buffer, concentrated in a centrivap to dryness and re-suspended in 30 uL dH2O. I took the concentration on a nanodrop three times for each sample. So, moment of truth, have I been wasting my time?

Hahahahaha…yeah, whoops. You don’t need any fancy statistics to tell you that there’s no difference between the two techniques. I did the stats for a laugh though (and scientific rigor), not even close to significant.

So my friends, challenge yourself early, feel stupid briefly, but save yourself tons of time in the long run and avoid passing down nonsense to the next generation.

9 thoughts on “Passing DNA through a silica column twice does not increase yield – A cautionary tale of cognitive bias”

  1. I presume it’s the same for gel extraction? Because I have issues with low yield using various kits.
    Also, could you test using warm elution buffer (70C) vs room temp?

    Thanks!
    Stavros

    1. Yep, I pass all the chaotropic solutions through my columns just once now, gel extractions, minipreps, PCR cleanups, don’t notice any loss in quantity out.

      Eluting in 100 uL of warm, weak TE (5mM Tris, 0.1 mM EDTA, pH 8) and concentrating in a centrivap is my go to method for high yields….honestly the centrivap is my best friend now, it’s become close to the nanodrop in terms of day to day use, so good.

      I have a followup post planned for questions like, does pre-warming make a difference? whats better, eluting twice or eluting in a higher volume etc etc

  2. And what about passing the final eluate through the column twice ? I usually do it when extracting DNA/RNA from low amount of initial material and think it actually increase yield.

    1. Will definitely test in the followup article, eluting twice with same volume, twice with two different volumes, one large volume, etc etc

      I mean, if your method works then its a win win, higher yield with same volume, I’d use that.

  3. Interesting… I have never passed DNA through a column twice, however I do often elute twice with fresh dH2O to have a “backup” sample or if I need a larger amount of DNA.

  4. For overnights, I usually grow them a little cooler (~25-30 C) so I know they’ll be exponential. Switching media can help (Wood et al., 2017). Danquah & Forde (2007) devised a media optimized for DNA yield, which they argue is ~5x cheaper than LB per mg of DNA. For the miniprep itself, I chill my neutralization buffer and always use hot (~80 C) water to elute my plasmids. Not sure about chilling the N3 buffer, but richer media and hot water usually get me good yields.

    1. I’m trying out the plasmid media buffer you mentioned, we’ll see how it goes when I get around the processing the pellets, thanks 🙂

      Don’t have too much trouble with yields, but if it means I have to maxiprep plasmids less often I’d love that.

  5. I just did the same, but different. I did two washes of the same column to two separate tubes. Second tube had about 20% of pDNA when compared to first one. Checked with qPCR. So, there is a loss of DNA on column.

  6. Also can you count you DNA loss on column vs. your initial concentration? Sorry I’m to lazy for a math… 😛

Leave a Reply

Your email address will not be published. Required fields are marked *