Infinite and modular molecular DNA spiral

This is a completely modular DNA model that can be extended infinitely to your own liking. Since DNA is a very complex molecule to print all together, I decided to design a model in which individual nucleotides can be printed and then hooked together afterwards. This model is DNA in its A-form (see https://en.wikipedia.org/wiki/A-DNA), which is the shape it takes when being dehydrated. I may make a similar model for B-form DNA in the future, please leave a message in the comments if you'd like that. 

The logic behind the printing and assembly is as follows:

• As usual, adenine (A) always pairs with thymine (T), held by two hydrogen bonds, and guanine (G) always pairs with cytosine (C), held by three hydrogen bonds. Other pairing is not possible.
• There are two strands, here indicated by strand 1 (S1) and strand 2 (S2).
• Looking at each individual strand, the phosphodiester bond between adjacent nucleotides (A/C/G/T) works with a stick-and-hole type connection. Within each strand, only use either S1 or S2 type bases.
• The hydrogen bonds between the two strands are also stick-and-hole type connections. The sticks are always on the S1 strand, and the holes are always on the S2 strand (this makes S1 and S2 nucleotides non-exchangeable)
• The double strand starts with a CG pair, indicated by S1-C-term and S2-G-term (term stands for terminated end), and it ends with a TA pair, indicated by S1-T-term and S2-A-term. So, you can make whatever DNA sequence, as long as you terminate the model with these two pairs.

Print settings, cleanup, and assembly

• In your slicer, enlarge the model to a size of your liking. I printed everything at 1000% scale.
• Rotate it so that the amount of support structures needed is minimized. For me, this usally meant that the ring structures were laying almost flat on the base-plate
• I used 100% infill density, PLA in “Light Marble” from Additive Heroes.
• I used a 4 mm brim and tree supports (everywhere), support overhang 50 deg, 90% support interface density, 10 deg tree angle, and 40 mm base. Together this created a continuous support structure, minimizing the risk of detachment of parts.
• The print time is around 6 to 7 hours per nucleotide, I usually printed two at the same time. Make sure to label them correctly after printing.
• Clean-up is a lot of work if you don't use a dissolvable support filament, but not more than usual. I used clippers, files, and a small sanding machine to remove the support structures and get a smooth enough finish.
• Finally, use instant glue to connect the nucleotides together. I recommend starting with making all the AT and CG pairs and then connecting them in your desired sequence. Since there was 0 tolerance in the design of the stick-and-hole connections, you may have to use a 3 mm drill to widen the holes and make them more circular. In my case, the sticks were also a bit coarse and needed a touch-up with a file. Before gluing, make sure the sticks can go in all the way. In the end, the glued connections were strong.

Mounting and display

For mounting the model in vertical direction, I used an 18 mm diameter steel pipe, that fit snug in the baseplate ("Baseplate.stl"). The baseplate was screwed on a piece of wood. I then designed clamps that tighten on the pipe using M3 nuts and bolts to create slots in which the model can be supported (see “Clamp for 18 mm rod.stl”. I used transparent fishing line to connect the model to the clamps, which worked very well.

How did I create this model?

1. Web 3DNA website was used to make a molecular DNA model in PDB format: http://web.x3dna.org/fibermodel/regseq 
2. Chimera was used to transform PDB into a Blender readable format (in this case x3D, see next point) and in a molecular ball and stick model that I liked, using the following model settings: 
   represent b+s 
   setattr M ballScale 0.35
   setattr M stickScale 1.3
   bondcolor grey 
   setattr p drawMode 1 
   vdwdefine 1.78 @o= 
   vdwdefine 1.8 @n= 
   vdwdefine 2.0 @c=
3. I used the MolPrint3D extension for Blender, made by Prof. Paul Paukstelis (https://chem.umd.edu/people/paul-paukstelis) to split up the model and create the stick-and-hole connections, which can be found at https://github.com/paukstelis/MolPrint . However, please note that you need an earlier version of Blender to get this working, 2.79b or earlier (see my comments and tips & tricks at https://github.com/paukstelis/MolPrint/issues/3 ). The final blend file is also supplied here.
4. Finally, the individual nucleotides were saved as STL files.