How much per-tile variation in sequence quality can be expected for a normal AVITI™ sequencing run?

Question:

How much per-tile variation in sequence quality (as analyzed by report programs like FastQC) can be expected for a normal AVITI sequencing run?

Answer:

Slight variation in tile quality is typical for an AVITI sequencing run because it provides continuous, rather than binned, quality scores. Generally, you can ignore this variation. However, if patterns are widespread, sustained, or correlated with performance issues, reach out to your Field Applications Scientist or Element Biosciences Support at support@elembio.com.

Summary

• Optical or chemical variability can cause slight differences in performance across a flow cell during sequencing
• Ignore any slight variations in tile quality which is typical for platforms that provide continuous quality scores
• Investigate any widespread or sustained tile quality loss

Considerations

What Is Tile Variation?
During sequencing, the AVITI Operating Software divides the flow cell into many tiles that are processed in parallel. Each tile can show slight differences in performance due to optical or chemical variability. This variation is visualized in tools like FastQC via the "Per Tile Sequence Quality" module.

When Is Variation Considered Normal?
The AVITI provides continuous (not binned) quality scores. When tile quality is visualized with tools like FastQC, variation can appear more prevalent in data with continuous Q-scores compared to data from platforms that bin quality scores (see Does the AVITI bin Q-scores? for details). Normal variation in tile quality often includes:

• A small number of tiles affected
• Drop in quality for only 1–2 cycles
• Isolated to specific coordinates or cycles
• No consistent spatial pattern across the flow cell

These are typically not cause for concern, especially if overall run metrics such as output, % passing filter, and quality scores are strong. Examples of normal variation for AVITI are shown in Figure 1A and 1B while Figure 1C shows an example of poor quality.

Figure 1. Examples of per tile sequence quality output from FastQC. (A) Good performance, (B) Okay performance, and (C) Poor performance.

When to Investigate Further:

• Persistent low-quality signals across many tiles or cycles
• Patterns of deviation that span large areas of the flow cell
• Signs of general flow cell overloading
• Recurrence of low quality in the same tile or position across multiple runs