Protocols Overview

FloodLAMP uses an assay chemistry developed by Brian Rabe and Prof. Connie Cepko at The Cepko Lab at Harvard Medical School. “Rabe Cepko,” as it’s known, uses an ultra cheap, streamlined front end (sample inactivation and RNA purification+concentration). This protocol has been chosen by many groups, from pharmaceutical companies to researchers working in the developing world. The same aspects that make it ideal for low-resource settings in Africa also make it ideal for America—it’s cheap, works well, and can be implemented by any basic chemistry or biology lab. The abstract and introduction of the recently published (9/8) peer reviewed PNAS paper are a great summary and highly recommended reading.

This assay was validated clinically in May by researchers at Mass General Hospital. In addition, the key amplification reagent (NEB Colorimetric LAMP Master Mix) is the same product used in the FDA-approved test by Color Genomics, who does about 5K tests per day in the SF Bay Area.

Below are our detailed protocols directly adapted from Rabe Cepko, who we thank for the original work and their ongoing collaboration with our group and others. These protocols are in worksheet format as we use them in the lab. We will continue to build out resources and content, publishing here and on other websites such as

Time permitting, we aim to produce several versions and formats for different audiences; in the meantime, we encourage interested people to sign up through the form on our CONTACT page.

Silica LAMP Assay Protocol

Silica LAMP Assay Protocol v4.1

This is the main protocol. It includes the glass milk (silica) based purification and LAMP amplification. On one hand, it is vanilla Rabe Cepko, following their protocol closely. On the other hand, it includes various workflow details which we’ve arrived at after extensive experimentation. 

Our current preferred version incorporates a transfer of the nucleic acid bound silica to the PCR tubes (where the LAMP reaction is carried out) in ethanol. The pellet resuspends easily in the 80% ethanol and—once in the PCR tubes—can be quickly dried on a heat block. Then they are ready for addition of the 1X LAMP Reaction Mix. We have used alternatives to this method, including an elution of the nucleic acid in water and also a resuspension of the pellet in water or 1X PBS, then addition to the LAMP reaction.

In addition to the LAMP Master Mix (NEB 1804), the assay uses 2 prepared solutions: a NaI Binding Solution and the Glass Milk (also called “prepared silica”). 

Inactivation Solution

Inactivation Solution Prep v1.0

The Inactivation Solution includes TCEP, EDTA, NaOH, and water. It should be made in a chemical fume hood, as the TCEP is caustic. It is very straightforward to make if starting with liquid-form TCEP. We have not made it with powder-form TCEP, which is about 30% cheaper at low-volume retail prices from Sigma-Aldrich. 

The prepared Inactivation Solution should be treated with care, as the TCEP is at 0.25M concentration and is caustic. It should be used only in a biosafety cabinet, hood, or behind a splash guard with proper PPE.

We have yet to do long-term stability studies; however, we think it’s fine stored at room temperature for at least a few weeks. If the yellow Binding Solution does not go clear upon addition to an inactivated sample or PBS+Inactivation Soln, suspect that the Inactivation Solution may have degraded, and repeat with another aliquot or freshly-made Inactivation Solution.

Inactivation Protocol

Inactivation Protocol v3.1

The inactivation protocol uses a chemical plus heat to break open cells (including virus, if present) and preserve the RNA. The inactivation can be done outside a lab setting, at the point of sample collection or drop-off. This is our preferred mode, and we refer to these as “inactivation stations”. Prior to inactivation, the samples may contain live, contagious virus, so it is crucial that personnel use proper PPE and safe handling procedures. Many current community screening efforts are performing non-lab based sample processing (for example, Dave O’Connor and Chris Mason), and we have followed their examples with the use of splash guards in addition to PPE.

After inactivation, samples should be refrigerated or stored on ice before same-day processing through the assay. If not using same day, inactivated samples should ideally be stored at -80C. Dave O’Connor and his group have reported sample degradation with storage at -20C.

Glass Milk (Prepared Silica)

Glass Milk Prep v1.0

This is a work-in-progress. We have not prepared the glass milk yet. We are still using two 50mL Falcon Tubes kindly sent to us by Brian Rabe. Each 50mL tube has enough for 10,000 individual 0.5mL samples. 

Binding Solution

Binding Solution Prep v1.1

The Binding Solution includes NaI, HCl, Triton-X100, and water. We have had lots of trouble with it, largely due to a mistake that has been corrected. In our hands currently, the Binding Solution is a pale yellow—a possibility mentioned in the Rabe Cepko paper. It does seem to become cloudy over time (days to weeks). Our current plan is to prepare weighed-out NaI and the other components premixed (stored at 4C) for easier prep, and then make fresh Binding Solution up at least weekly.

We have moved to storing the Binding Solution in 5mL amber screw cap tubes for development runs (2.25mL for a single strip of 8 samples), with these in an opaque container. This makes it convenient to add glass milk and use it as a master mix (as suggested by Rabe Cepko). Care must be taken to keep the silica resuspended prior to pipetting for the addition to the sample, hence the mixing by pipetting on every draw.

Equipment Notes

We utilize heat blocks and have eliminated water baths from our protocol because 1) danger of steam burns and 2) potential for contamination from the water. We also avoid ice baths because of contamination. However water baths and ice can be utilized with adequate precautions, such as additives or additional ethanol spritzing. 

We’ve selected a low-cost centrifuge (SpinPlus $276) for our 4 current inactivation stations. Each centrifuge holds six 15mL or 5mL tubes. Based on processing one batch of 6 tubes every 5min, this process takes about 20 minutes total per batch of sample tubes, which can be pipelined through the heat, cooling and centrifugation steps.