3.6 Infrastructure considerations
Establishing a WGS facility requires you to consider whether the existing infrastructure you have is suitable.
Which of the four stages of WGS could potentially require dedicated infrastructure and high capital investment?
Answer
The most likely are the wet-lab processes (a laboratory facility for sample culture, DNA isolation, quantitation and library preparation), DNA sequencing (a laboratory facility and instrumentation), and data analysis.
As part of the planning process, suitable facilities will need to be put in place for these stages.
In addition to laboratory space, when establishing a WGS centre it’s essential to ensure reliable power and internet access to avoid costly disruptions during sequencing and analysis. If power is interrupted during an Illumina run, the sequencing process usually cannot be resumed, often resulting in the loss of reagents and the need to restart the entire run. By contrast, Oxford Nanopore Technologies (ONT) runs can sometimes be paused and resumed, depending on the set-up and software version. Continuous power is also important for keeping reagents and samples refrigerated or frozen.
Similarly, during the data-analysis phase, internet connectivity plays a key role. Cloud-based platforms such as BaseSpace or Epi2Me depend on uninterrupted internet access to function properly: if the connection is lost, the analysis may stall or fail. However, local tools and offline workflows (e.g. command-line pipelines) can typically continue running without internet access, provided the required data and tools are already downloaded.
Another critical infrastructure consideration is data storage (Figure 10). Especially when it is done at scale, sequencing generates large volumes of data that must be securely stored and managed.
As an example of the storage requirements, sequencing a single E. coli isolate typically produces around 250–500 MB of data with short-read machines like Illumina, and approximately 150–300 MB with long-read machines like ONT. Even though ONT reads are much longer, fewer are needed to sequence a genome, so the total amount of data generated can be smaller than with short-read technologies.
Both machines generate additional files during downstream analysis, such as assemblies and resistance-gene reports, that can quickly consume terabytes of space when processing large numbers of samples. Institutions may choose between cloud storage (for flexibility and scalability) or local servers (for faster access and better control – especially in low-bandwidth environments).
You’ll also need to consider the computational resources required for data analysis. Web-based tools like EPI2ME, BaseClear and PathogenWatch are user-friendly and require minimal local computing power, making them ideal for beginners or low-resource settings. However, more advanced or large-scale tasks (e.g. genome assembly, variant detection or AMR surveillance) require a high-performance computer with a strong processor, sufficient RAM and ample hard drive space.
Careful planning across power supply, internet connectivity, data storage and computing capacity is essential to build efficient, scalable and sustainable WGS operations, particularly in resource-limited settings.
Activity 7: Reflection on infrastructure needs for establishing a WGS lab
Look at the questions in Table 4 to help you reflect on your laboratory’s readiness for introducing WGS. (You may want to use the space below to make notes.)
| Question | Reflection prompts |
|---|---|
| Do I have consistent power and internet access? | If not, can I supplement with back-up solutions like uninterruptible power supplies (UPS) or generators? Can I install a redundant internet line or use local/offline tools when needed? |
| Where will I store my sequence and analysis data? | Are there any legal or policy restrictions on using cloud storage in my country or region? Do I need to invest in local servers with back-up? |
| How intensive will my data analysis be? | Will my current computer be sufficient or do I need a high-performance workstation with more RAM, storage and CPU cores? |
| Am I prepared to share data with collaborators or global databases? | Do I understand the relevant data-sharing policies (e.g. for ENA, GISAID or public health databases)? Is ethical or patient consent required? |
| Do I need a single, centralised WGS centre or several smaller centres? | Would several smaller, decentralised centres improve turnaround time and sample access, or is a central hub more efficient for coordination and cost? |
| Do I have a trained bioinformatician on staff? | If not, can I hire, train or partner with an external bioinformatics provider? Should I consider using cloud-based plug-and-play tools as a short-term solution? Should I send my data to an external provider for analysis? |
| Do I have a well equipped and functioning microbiology laboratory? | If not, can I invest in improving basic microbiology capacity, including proper sample handling, species identification and AST? Can I partner with existing reference labs or central facilities to ensure high-quality isolates are selected for sequencing? |
| How will I select and collect samples for sequencing? | Do I have a clear sampling strategy for routine surveillance, outbreak response or special studies? Are metadata and tracking systems in place? |
| What is more important for me and my work: high throughput and cost-effectiveness, or accuracy and higher-quality data? | If high throughput and cost-effectiveness are more important, short-read sequencing might be more suitable. Small-volume laboratories may find that lower-throughput machines are more useful because they won’t require you to wait to fill a high-volume machine with samples. |
Discussion
You’ve done well to thoughtfully assess your laboratory’s infrastructure and readiness for WGS implementation. If you have difficulty identifying potential gaps or resources, take the time now to consult with colleagues who are expert in these areas. As you move forward, continue to align your infrastructure planning with your specific goals.
Having learned about the practicalities of establishing a WGS facility, in the next section the focus moves onto examining how a laboratory might be integrated into a wider surveillance network.
3.5 Data-sharing considerations
