Flattening the Curve with Nextflow:

Building a Scalable and Reproducible Bioinformatic Workflow for MITNANEX

By Loka’s Bioinformatics team: Jelena Pejovic, Bioinformatics Engineer; José Ribón, Technical Project Manager; Andrés Sacre, Data Engineer; Federico Rueda, Data Engineer; Nicolás Franco, Data Engineer; Javier Dominguez, Data Engineer; Juliana Silva, Marketing Designer

Loka’s participation in the recent nf-core hackathon in Medellín, Colombia, culminated in the successful development of MITNANEX, a pipeline designed to extract and assemble mitochondrial genomes, identify and annotate variants and facilitate comprehensive mitochondrial genome annotation. By leveraging nf-core modules and the workflow orchestration tool Nextflow, our team built a solution to explore mtDNA variants potentially linked to specific traits in the human organism.

The results of our hackathon efforts streamline MITNANEX’s path to answering its key research question:

Do certain mitochondrial variants correlate with exceptional longevity in humans?

The nf-core Hackathon

The hackathon was part of a globally coordinated initiative led by nf-core and Seqera designed to empower bioinformaticians, developers and researchers to build modular workflows using Nextflow. Loka joined forces with EAFIT University to bring this opportunity to Medellín, opening a space where students and professionals could co-create, test and contribute directly to the nf-core ecosystem.

A modular approach allowed multidisciplinary subteams to focus on specific components that were then seamlessly integrated, fostering valuable knowledge sharing between participants of diverse backgrounds. Our collaboration led to proposing MITNANEX as a future nf-core module, with a roadmap defined by Loka and EAFIT. It demonstrated how synergies between academia and industry can lay the foundations for new research and accelerate innovation in the bioinformatics community.

The Nextflow & nf-core Framework

Nextflow has changed research and development in bioinformatics and revolutionized the way computational pipelines are built, shared and executed by enabling smooth reproducibility and scalability across platforms. It supports scalability, reusability, portability and efficiency while processing large amounts of data across different infrastructures and tools.

This is where nf-core comes in: a collection of well-maintained, community-curated pipelines and modules that follow best practices and are designed for portability and reuse out of the box. With nf-core modules, it’s easier to plug in standardized components without the need to build everything from scratch, saving time and potential error. It also means that the same code can run across different environments with easy setup.

Users benefit from workflows that are not only reproducible and portable but also easy to implement and well-documented. Developers, on the other hand, can utilize nf-core’s readymade templates and tools to simplify the creation of robust Nextflow pipelines.

This is possible thanks to nf-core’s vibrant and collaborative community.

Introducing MITNANEX

MITNANEX (MITochondrial NANopore reads EXtractor) arose from the observation that many cell processes tied to aging, from oxidative stress to metabolic efficiency, have been traced back to the mitochondria. By analyzing the mitochondrial genomes of people aged 100 and above, researchers aim to uncover shared mutations, deletions or even epigenetic signatures that could explain their extended cellular vitality.

Processing mitochondrial genomes involves several steps, especially when working with Oxford Nanopore Technologies (ONT) data. Oxford Nanopore reads present unique challenges due to their higher error rates compared to Illumina short reads, which can lead to inaccuracies in base calling. However, their longer read lengths allow for the assembly of more complex genomes, such as mitochondria, in a single pass, making them ideal for de novo assembly where no reference genome is available or where reference bias must be avoided, despite the need for specialized error-correction tools. Currently there are few pipelines specifically designed to process mitochondrial genomes directly from whole genome sequencing (WGS) data.

MITNANEX Case Study

MITNANEX is a pipeline designed to de novo extract mitochondrial reads from whole genome sequencing (WGS) data generated by Oxford Nanopore, without the need for seeds or references. It produces a draft assembly of the mitochondrial genome using the assembler Flye.

Overall, MITNANEX streamlines mitochondrial genome recovery with these efficient and integrated steps:

1. Preprocessing & Filtering

• Chopper: Processes FASTQ data to remove unwanted reads (filtering).

2. Alignment or Assembly (Two paths)

a) Reference-based alignment:

• Clean reads → minimap2 (aligned to mitochondrial reference)
• Aligned reads → samtools bam2fq to extract mitochondrial reads
• Extracted reads → MetaFlye for mitochondrial de novo assembly

b) Direct de novo path:

• Clean reads → De-novo assembly → filtered by seqkit → De-novo mitochondria

3. Annotation

• If a reference GFF is available: Liftoff can transfer annotations
• If not: Proceed with de novo annotation on assembled genome

4. Read Mapping & Variant Calling

• Assembled genome → minimap2 for read mapping
• GATK Mutect2 → variant calling
• GATK FilterMutectCalls → variant filtering
• Generates VCF file

5. VCF Post-processing

• Annotate with Bedtools and Mitomaster

So far, MITNANEX has been successfully tested on fungal mitogenomes, achieving near-perfect results: 100% query coverage and over 99% identity when validated via blast against NCBI references. The Loka team is planning more extensive benchmarking but these early results suggest strong performance, especially considering the challenges of working with noisy Oxford Nanopore data. Because it’s built with Nextflow, the pipeline is easy to run, share and adapt to new datasets or environments, making it a solid foundation for future development. Compared to more manual or ad hoc approaches, MITNANEX offers a solution that’s more reproducible and modular and can be quickly extended for other organisms or research goals.

Next Steps

Looking ahead, we plan to refine MITNANEX into an official nf-core module, share it with the broader scientific audience and make sure our work is cycling back to the nf-core and Nextflow community. By embracing the spirit of open source — creating something valuable together and letting it evolve as others build upon it — we hope to accelerate discoveries and unlock new insights into human aging.