We have DNA sequences!

It’s been two years since we set out to generate DNA sequence data from the seed samples from the Community Research Network, and we have now acquired DNA sequences for all samples sent to me between spring 2013 and fall of 2015! These sequences are the key to identifying the bacterial and fungal species inherited of the seeds that we save. It is not an easy process. First, it requires sterilizing the surface of the seeds, and then grinding them in a sterile environment, to prevent non-seed-associated microorganisms from contaminating the sample. Then, we extracted the DNA from the ground-up seeds. The most difficult process has been obtaining sufficient microbial DNA product, which is obscured in an ocean of plant DNA. The low-quantity microbial DNA was amplified to sufficient quantities such that we could obtain microbial DNA sequences, rather than those of the plant.

The process took many hours of troubleshooting and moving tiny liquid volumes between many plastic tubes. Now, it will take many more hours on the computer to pour over millions of DNA samples, and begin to disentangle the important factors that determine which microbes become inherited each new generation. If you sent us samples, you may receive call in the coming weeks, asking for some more details of your farming techniques, land history, seed source, etc. You should expect to see more regular updates on this website as the DNA sequence data become more organized and clear.

Meanwhile, we would like to do more sampling! We have streamlined the laboratory process, and are eager to receive more samples for DNA sequencing in the future. The goal is to keep growing the database, in terms of gaining more research participants and more corn varieties, but also getting more years of corn seed harvest from the same community research members. If you are interested in sending seeds, please write an email to lnebert at uoregon.edu for more information.


Each seed sample was reduced to a tiny pool of DNA sitting at the bottom of a 96-well plate.


576 samples containing amplified microbial DNA from six 96-well plates were pooled together.


This tiny volume of liquid contained millions of microbial DNA sequences from 288 samples, with each sample containing its own DNA barcode so it could be linked back to the sample at a later time. The final volume was 40 ul (0.4% of 1 milliliter).

Screenshot from 2016-02-04 15:25:46

Here is an example of a few bacterial sequences that were identified from the seeds. Note the number of nitrogen-fixing bacteria! Corn and other grasses can team with nitrogen fixing bacteria to provide significant amounts of their nitrogen needs, contrary to common knowledge that this process is only found in legumes in agriculture.

Screenshot from 2016-02-04 14:47:09

Here is an example of a Principle Component Analysis (PCoA) plot. Each dot is a seed sample, and the distance between dots is the differences of their entire microbial communities. They are currently colored by ‘type’ such as flint, dent, flour, popcorn and sweet corn. Type does not seem to be a good indicator of seedborne microbes, so we are working to find which factors are most important in determining which microbes are inherited.