Hans Zauner

**This week in *GigaScience*, Hasindu Gamaarachchi and colleagues
published a paper on their [HARU solution for selective
sequencing](https://doi.org/10.1093/gigascience/giad046), to be used
alongside the handheld (["tricorder"
like](http://gigasciencejournal.com/blog/iphone-genomics-beaming-us-up-to-the-tricorder-era/))
Nanopore MinION device**.

With the MinION, it is possible to kick out all non-target DNA strands
before they are processed through the device's pores, thus selectively
sequencing specific target regions of the genome. This is a clever way
to save time and resources, in principle. However, the computational
demands for this method are challenging and usually require an expensive
cluster infrastructure. 

Having a tiny portable sequencer is nice. But the miniaturisation of the
sequencing hardware is not satisfactory if you need bulky and expensive
computers to do the real-time analysis steps that are necessary during a
selective sequencing run.

As Dr.  Gamaarachchi  explains in our
[Q&A](http://gigasciencejournal.com/blog/tag/qa/) below, the authors'
motivation therefore was to realise a portable, open source computer
system that can efficiently perform the analysis tasks of selective
sequencing; in a tiny \$300 device that is "*two times faster than a
30,000 \$ 36-core server, at a fraction of power consumption*".  The
proof-of-concept solution, called HARU, is suitable for small genomes
such as viruses or bacteria, but the authors hope that their work will
lead to further improvements down the line: "*Our dream would be to one
day have a chip within a nanopore sequencer that can handle the
computational requirement of selective sequencing directly.*",
Gamaarachchi concludes.

## **Author Q&A with Hasindu Gamaarachchi** {#author-qa-with-hasindu-gamaarachchi .wp-block-heading}

### ***"What is selective nanopore sequencing and how does it work"?*** {#what-is-selective-nanopore-sequencing-and-how-does-it-work .wp-block-heading}

Selective nanopore sequencing is a nascent method to perform rapid
genome analysis that enables the sequencing of specific genomic "regions
of interest" rather than the entire genome. This programmable approach
is free from biochemical processes, making it highly flexible and
cost-effective.

::: {.wp-block-image}
<figure class="alignleft size-medium is-resized">
<img
src="http://gigasciencejournal.com/blog/wp-content/uploads/2023/07/author-300x282.jpg"
class="wp-image-5022" loading="lazy" decoding="async"
srcset="http://gigasciencejournal.com/blog/wp-content/uploads/2023/07/author-300x282.jpg 300w, http://gigasciencejournal.com/blog/wp-content/uploads/2023/07/author.jpg 422w"
sizes="(max-width: 225px) 100vw, 225px" width="225" height="212" />
</figure>
:::

Nanopore sequencers utilise an array of tiny pores, known as nanopores,
to measure and record fluctuations in ionic current as DNA/RNA strands
pass through them. If the region is irrelevant to the given research
problem, the voltage across the nanopore can be reversed to eject the
DNA/RNA strand from the pore. This approach optimises sequencing time
and consumable resources (flowcells) by focusing primarily on the
regions of interest.

### ***"What are application examples for selective sequencing?"*** {#what-are-application-examples-for-selective-sequencing .wp-block-heading}

Selective sequencing can reduce the time and cost of sequencing in
applications such as genetic disease diagnosis, cancer detection, the
surveillance of viruses (e.g. SARS-CoV-2) and other pathogens, and
sequencing low abundance species in an environmental sample containing
thousands of species. For example, it has been recently demonstrated
that nanopore selective sequencing can be used to create a [single test
for over 50 genetic diseases that will cut diagnosis cost and
time.](https://www.science.org/doi/10.1126/sciadv.abm5386)

### ***"What are the technical and computational challenges that come with this method?"*** {#what-are-the-technical-and-computational-challenges-that-come-with-this-method .wp-block-heading}

The real-time analysis of the data streamed from the sequencer is
computationally demanding. The rapid analysis of data from hundreds of
nanopores is crucial to promptly identify and eject unnecessary DNA/RNA
strands. Any delay in this processing would undermine the effectiveness
of selective sequencing. While nanopore sequencers like the Oxford
Nanopore MinION are small portable devices comparable in size to a
mobile phone and powered via USB, the computer required for performing
selective sequencing is typically the size of a non-portable small
fridge, consuming power that is hundreds of times greater than the
MinION.

::: {.wp-block-image}
<figure class="aligncenter size-full is-resized">
<img
src="http://gigasciencejournal.com/blog/wp-content/uploads/2023/06/HARU.jpg"
class="wp-image-4985" loading="lazy" decoding="async"
srcset="http://gigasciencejournal.com/blog/wp-content/uploads/2023/06/HARU.jpg 815w, http://gigasciencejournal.com/blog/wp-content/uploads/2023/06/HARU-300x189.jpg 300w, http://gigasciencejournal.com/blog/wp-content/uploads/2023/06/HARU-768x484.jpg 768w"
sizes="(max-width: 611px) 100vw, 611px" width="611" height="386" />
</figure>
:::

### ***"What was your motivation to work on this problem"?*** {#what-was-your-motivation-to-work-on-this-problem .wp-block-heading}

The usefulness of the nanopore MinION as a portable device is negated if
it depends on a computer which is non-portable. Although nanopore
selective sequencing has the potential to revolutionise diagnostic
techniques in the future, it necessitates portable computational
solutions to be utilised in settings such as clinics. Our motivation was
to realise a portable computer system that can efficiently do the
analysis side of the selective sequencing, which is synergistic with the
size of the MinION.

### ***"What is the main advance of your solution, HARU?"*** {#what-is-the-main-advance-of-your-solution-haru .wp-block-heading}

HARU is a complete system that works on off-the-shelf devices, as
opposed to being a conceptual work limited to simulation. In other
words, it is a tangible prototype that actually runs. The full proposed
design is available open-source. We have pre-compiled binaries and
step-by-step instructions on building the overarching HARU system from
scratch. One can simply purchase Xilinx's Kria AI Starter Kit, flash the
device, and execute HARU.

Advantages of HARU include portability, low cost, high-processing
throughput, low power consumption and scalability. HARU can also be used
as a framework for other future work intending to explore acceleration
for selective sequencing on hardware.

### ***"What about the costs? How much would it cost to implement the software/hardware setup as presented in the article?"*** {#what-about-the-costs-how-much-would-it-cost-to-implement-the-softwarehardware-setup-as-presented-in-the-article .wp-block-heading}

The only cost to replicate the HARU system is the target board itself.
The software is free and open-source. We target Xilinx's Kria AI Starter
Kit, which is currently retailed at \$300. As we have shown in the
paper, for selective sequencing based on the algorithm called
subsequence Dynamic Time Warping, this tiny \$300 device is 2X faster
than a 30,000\$ 36-core server, at a fraction of power consumption.

### ***"Is this a one-off proof of concept, or are there plans to develop this system for routine use?"*** {#is-this-a-one-off-proof-of-concept-or-are-there-plans-to-develop-this-system-for-routine-use .wp-block-heading}

While HARU is currently a proof-of-concept, we envision continuing
development and integration of this system into the nanopore sequencing
ecosystem for routine use. However, this is a long journey as a lot of
development has to be conducted to interface with the nanopore sequence
acquisition and control software (closed source) called MinKNOW. Also,
HARU is currently only suitable for small genomes such as viruses or
bacteria and is impractical for large genomes like the human genome. We
have thoughts to combine HARU with fast index-based mapping methods to
tackle the problem with such large genomes. Open-sourced HARU can be
used as a stepping stone by other researchers who would like to explore
methods of hardware acceleration for selective sequencing.  Our dream
would be to one day have a chip within a nanopore sequencer that can
handle the computational requirement of selective sequencing directly.

#### References {#references .wp-block-heading}

**Efficient real-time selective genome sequencing on
resource-constrained devices**  
Po Jui Shih, Hassaan Saadat, Sri Parameswaran, Hasindu Gamaarachchi  
GigaScience, Volume 12, 2023, giad046,
<https://doi.org/10.1093/gigascience/giad046>

Source code for the HARU accelerator: <https://github.com/beebdev/HARU>

The post [Selective sequencing on a shoestring: the \$300 HARU
system](http://gigasciencejournal.com/blog/selective-sequencing-on-a-shoestring-the-300-haru-system/){rel="nofollow"}
appeared first on
[GigaBlog](http://gigasciencejournal.com/blog){rel="nofollow"}.


<strong>
 This week in
 <em>
  GigaScience
 </em>
 , Hasindu Gamaarachchi and colleagues published a paper on their HARU solution for selective sequencing, to be used alongside the handheld (“tricorder” like) Nanopore MinION device
</strong>
.   With the MinION, it is possible to kick out all non-target DNA strands before they are processed through the device’s pores, thus selectively sequencing specific target regions of the genome.

Selective sequencing on a shoestring: the $300 HARU system

Maëlle Salmon

As you might remember from [my blog post about
`ropenaq`](/blog/2017/02/21/ropenaq), I work as a data manager and
statistician for an [epidemiology project called
CHAI](http://www.chaiproject.org/) for Cardio-vascular health effects of
air pollution in Telangana, India. One of our interests in CHAI is
determining exposure, and sources of exposure, to PM2.5 which are very
small particles in the air that have diverse adverse health effects. You
can find more details about CHAI [in our recently published protocol
paper](https://www.ncbi.nlm.nih.gov/pubmed/28606699). In this blog post
that partly corresponds to the content of [my useR! 2017 lightning
talk](https://sched.co/AxrS), I'll present a package we wrote for
dealing with the output of a scientific device, which might remind you
of similar issues in your experimental work.

## Why write the rtimicropem package?

![](/assets/blog-images/2017-08-29-rtimicropem/Allequipment_Frontview_cropped.jpg)

Part of the CHAI project is a panel study involving about 40 people
wearing several devices, as you see above. The devices include a GPS, an
accelerometer, a wearable camera, and a PM2.5 monitor outputting
time-resolved data (the grey box on the left). Basically, with this
device, the RTI MicroPEM, we get one PM2.5 exposure value every 10
seconds. This is quite exciting, right? Except that we have two main
issues with it...

![](/assets/blog-images/2017-08-29-rtimicropem/screenshot_output2.png)

First of all, the output of the device, a file with a ".csv" extension
corresponding to a session of measurements, in our case 24 hours of
measurements, is not really a csv. The header contains information about
settings of the device for that session, and then comes the actual table
with measurements.

Second, since the RTI MicroPEMs are nice devices but also a
work-in-progress, we had some problems with the data, such as negative
relative humidity. Because of these issues, we decided to write an R
package whose three goals were to:

- Transform the output of the device into something more usable.
- Allow the exploration of individual files after a day in the field.
- Document our data cleaning process.

We chose R because everything else in our project, well data processing,
documentation and analysis, was to be implemented in R, and because we
wanted other teams to be able to use our package.

## Features of rtimicropem: transform, explore and learn about data cleaning

First things first, our package lives
[here](https://github.com/ropensci/rtimicropem) and is [on
CRAN](https://cran.r-project.org/web/packages/rtimicropem/index.html).
It has a [nice documentation
website](https://docs.ropensci.org/rtimicropem/) thanks to
[`pkgdown`](https://github.com/hadley/pkgdown).

### Transform and explore single files

In `rtimicropem` after the use of the `convert_output` function, one
gets an object of the R6 class `micropem` class. Its fields include the
settings and measurements as two `data.frames`, and it has methods such
as `summary` and `plot` for which you see the static output below (no
unit on this exploratory plot).

![](/assets/blog-images/2017-08-29-rtimicropem/plotexample.png)

The plot method can also outputs an interactive graph thanks to
[`rbokeh`](https://hafen.github.io/rbokeh/).

While these methods can be quite helpful to explore single files as an R
user, they don't help non R users a lot. Because we wanted members of
our team working in the field to be able to explore and check files with
no R knowledge, we created a Shiny app that allows to upload individual
files and then look at different tabs, including one with a plot, one
with the summary of measurements, etc. This way, it was easy to spot a
device failure for instance, and to plan a new measurement session with
the corresponding participant.

### Transform a bunch of files

At the end of the CHAI data collection, we had more than 250 MicroPEM
files. In order to prepare them for further processing we wrote the
`batch_convert` function that saves the content of any number of
MicroPEM files as two (real!) csv, one with the measurements, one with
the settings.

## Learn about data cleaning

As mentioned previously, we experienced issues with MicroPEM data
quality. Although we had heard other teams complain of similar problems,
in the literature there were very few details about data cleaning. We
decided to gather information from other teams and the manufacturer and
to document our own decisions, e.g. remove entire files based on some
criteria, in a [vignette of the
package](https://docs.ropensci.org/rtimicropem/articles/chai_data_cleaning.html).
This is our transparent answer to the question "What was your experience
with MicroPEMs?" which we get often enough from other scientists
interested in PM2.5 exposure.

## Place of rtimicropem in the R package ecosystem

When preparing `rtimicropem` submission to rOpenSci, I started wondering
whether one would like to have one R package for each scientific device
out there. In our case, having the weird output to deal with, and the
lack of a central data issues documentation place, were enough of a
motivation. But maybe one could hope that manufacturers of scientific
devices would focus a bit more on making the output format
analysis-friendly, and that the open documentation of data issues would
be language-agnostic and managed by the manufacturers themselves. In the
meantime, we're quite proud to have taken the time to create and share
our experience with `rtimicropem`, and have already heard back from a
few users, including one who found the package via googling "RTI
MicroPEM data"! Another argument I in particular have to write R
packages for dealing with scientific data is that it might motivate
people to learn R, but this is maybe a bit evil.

What about the place of `rtimicropem` in the rOpenSci package
collection? After [very useful
reviews](https://github.com/ropensci/software-review/issues/126) by
[Lucy D'Agostino McGowan](https://github.com/LucyMcGowan) and [Kara
Woo](https://github.com/karawoo) our package got onboarded which we were
really thankful for and happy about. Another package I can think off the
top of my head to deal with the output of a scientific tool is
[`plater`](/blog/2017/02/06/plater-blog-post). Let me switch roles from
CHAI team member to rOpenSci onboarding co-editor here and do some
advertisement... Such packages are unlikely to become the new `ggplot2`
but their specialization doesn't make them less useful and they fit very
well in the "data extraction" of the [onboarding
categories](https://devguide.ropensci.org/policies.html#package-categories).
So if you have written such a package, please consider submitting it!
It'll get better thanks to review and might get more publicity as part
of a larger software ecosystem. For the `rtimicropem` submission we took
advantage of the joint submission process of rOpenSci and the Journal of
Open Source Software, JOSS, so now our piece of software has [its JOSS
paper with a
DOI](https://joss.theoj.org/papers/7ead5e9a445da3e885d99247c5d6e58e).
And hopefully, having more submissions of packages for scientific
hardware might inspire R users to package up the code they wrote to use
the output of their scientific tools!


As you might remember from my blog post about ropenaq, I work as a data manager and statistician for an epidemiology project called CHAI for Cardio-vascular health effects of air pollution in Telangana, India. One of our interests in CHAI is determining exposure, and sources of exposure, to PM2.5 which are very small particles in the air that have diverse adverse health effects.

Rogue Scholar Posts

Selective sequencing on a shoestring: the $300 HARU system

rtimicropem: Using an R package as platform for harmonized cleaning of data from RTI MicroPEM air quality sensors