Nader's Powerpoint (Converted to PDF)

This section of the notes by Jonathan Magasin. They are keyed by title to Nader's slides.

Sample preparation and creating a good libary are key to getting good reads. (At $5-10K
in reagents per run, it matters!) Amplification is very important. We need both enough
fragments but also to be cautious not to create a polyclonal sample.

Moving to magnetic affinity resins rather than silica matrix which is currently popular.

Lysis took around 2hrs for banana slug.

Extraction: Phenol is not used so much anymore. It is noxious and gives a nasty headache.

Purification: Washing with higher salt solution and ethanol [confirm].

The matrix principle is the same whether it is bound or not.

Extraction: Usually do two elution rounds with warm water.

Comments on variability:

• Silica based [matrices] sometimes lose up to 80% of material.
• Magnetic based, lose not more than 15-20%.
• [Frustration with Nanodrop due to ten-fold variation.]
• Pipetting can lead to 50% variation, e.g. due to material on the tip.
• With Agilent Bioanalyzer they see 20% variation.

The first slug they caputured had bacteria covering its skin and so they could not use it.

The lanes smeared in a 2hr incubation. Maybe because AT-rich genome (but compared to
Helicobacter pylori?!…) If they left it overnight, there was no DNA left!

Focuses a beam (Adapted Focused Acoustics) to shear the DNA. The water bath keeps the
transducer cool (12°C). Very important as this is $100K equipment. Covaris advantage: no
sequence bias [of reagents] because nothing added to the sample (just the beam required
for shearing).

Effectiveness of cheap sample vessels discussed.

DNA digestion depends on the beam. You can tune it for desired size range, e.g. 300-600
bases for Illumina. You can get up to 1000 bases with Covaris. Note however that it cannot
get you the the roughly 3000 base fragments needed for paired ends.

With Covaris the beam can damage the DNA bases and sugar resulting in non-functional
template. Maybe losing 30% of your mycelles due to this. A new assay being investigated
to address this problem.

The current banana slug SOLiD run used templates generated by Covaris.

The same thing from your local pharmacy is $5. Even if you pay $25 from …. this is still
cheaper and simpler than Covaris. The fragment size is optimized for 454 and Illumina and
cannot be used for SOLiD. One issue is that DNA can bind to plastic and the Nebulizer has a
large surface area, so you're loosing at least half your material.

We are eluting from a gel. The volume is very large so use magnetic beads to concentrate.
They're switching to HPLC for more consistent collection/elution.

You cannot afford to lose 50% of cells if rare (e.g. T-cells). Imagine if you're using
a single cell. You can't afford to lose 30-40% of genomic material.

End repair (polishing) because you need to make blunt ends for adapter ligation.

To avoid bias, we don't want too many PCR amplifications. Check every 2 cycles for a
visible band and then stop PCR.

They're always seeing some small fragments regardless of what part of the gel cut (5%?)
and this causes problems in emulsion PCR.

Requires fewer steps and less input. Built-in die (FAM standard, attached to fragments)
so you can analyze the amount of material using a fluorometer. The yield is much less
than standard library preparation.

One bead might have a million fragments, another less, thus the 4-base key for normalization.

The 10-base barcode (MID) allows you to uniquely assign IDs to samples, so you can mix
samples and run them in the same channel or the same physical area.

Recall that SOLiD uses very short fragments (50 bases) so you don't want your first bases,
which are the highest quality, to be the barcode. Same for Illumina where the barcode is
separate from the fragment.

Area under curve is the amount of material. The two spikes are the ladders.

TissueLyser shakes the emulsion at a specified frequency. Higher frequency for smaller beads.
One micrometer beads require shaking much faster than for the 454, but it also depends on the
volume, tube size, shape and kind of oil.

Sequencing requires the full amount of reagents even if you do a partial run. To justify cost,
it's better to do full runs.

Getting the best yields: Too large mycelles and you get multiple beads or fragments in them.
However you need the mycelles large enough so there is buffer for the PCR reaction. It is a
balance.

Oil is very viscous and you so have to be careful not to break your mycelles when dispensing
them onto PCR plates. Nader said this is more art than science:)

Slide with pipetting picture insert.

The plastic caps have much better yield than …

Slide with mycelles picture insert.

After PCR, take a few pictures to see what your mycelles look like.

You have to wash with butanol to separate the beads from oil. Any remaining oil on the beads
interferes with sequencing reaction. This step is time consuming, at least 2 hours and boring.
Nader is trying to automate this (robot!).

Now you need to capture just the beads with amplified DNA. If using magnetic beads, only those
with amplification will capture the enrichment beads and get stuck to magnet.

Can use a FACs machine to sort for mono/polyclonality. (It's fast enough.) It sorts based on
color and so can separate out beads with a single color (monoclonal) from mixed color
(polyclonal). The dye is attached just after the primer based on 2 bases.

DNA is sheared from the beads with repeated microfluidics. Enriching for beads with good
signal is important especially if you have long templates.

No bubbles! When you lift the pipette it can leave a bubble that will move about during
sequencing. To avoid this you have to continue injecting material as you remove the
pipette. (Art!)

We have Titanium in house. Our banana slug reads were all done on Titanium.