> For the complete documentation index, see [llms.txt](https://docs.omics.ai/products/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.omics.ai/products/workbench/variant-interpreter/getting-started/the-summary-tab.md).

# The Summary tab & priority findings

When you open a sample for review, the Variant Interpreter lands on the **Summary** tab. This is the orientation surface for the case. It brings together:

* **Priority findings**: the variants most likely to explain the phenotype, ranked by strength of evidence.
* **Haplotype phase**: whether reported pairs of small variants in the same gene are in cis or in trans.
* **Run quality**: coverage, phasing, and dropout at clinical loci, so you know how far to trust the calls.
* **Technical detail**: the same quality measures per sample.

If the sample's files are ingested but no analysis has run yet, the tab shows a **Ready to process this sample** card instead.

## Priority findings

Priority findings are drawn from every variant type in the case and gathered into one ranked list. The header shows how many findings there are. Its subline tells you what the ranking used, including how many HPO terms the sample has and whether it is a trio.

### What qualifies as a priority finding

| Variant type                                | Qualifies when                                                     |
| ------------------------------------------- | ------------------------------------------------------------------ |
| **Small variants**                          | Exomiser's ACMG classification is Pathogenic or Likely Pathogenic. |
| **Structural variants** and **Copy number** | The automated classification is Pathogenic or Likely Pathogenic.   |
| **Tandem repeats**                          | The locus is an LPS outlier.                                       |
| **Methylation**                             | The region is strongly hypomethylated or hypermethylated.          |
| **Paralogs**                                | Not included in priority findings.                                 |

A variant you have classified as Pathogenic or Likely Pathogenic also qualifies, whatever its automated classification.

{% hint style="info" %}
Tandem-repeat priority findings count **LPS outliers only**. The Tandem Repeats tab opens on a broader default view (clinically relevant loci *and* outliers), so the two counts can differ. This is intentional. See [Tandem repeats](/products/workbench/variant-interpreter/reviewing-variants/tandem-repeats.md).
{% endhint %}

### Evidence tiers

By default, findings are sorted by **Evidence** into three sections: **Strong evidence**, **Moderate evidence**, and **Weak evidence**. A finding's tier reflects how much of the following evidence supports it:

* **Phenotype match**: how well the gene matches the sample's HPO terms.
* **Rarity**: how rare the variant is in control populations.
* **Constraint**: how intolerant the gene is to loss of function or dosage change.
* **Gene–disease validity**: whether the gene has an established disease association.
* **Inheritance**: whether the variant is de novo or inherited, where a trio makes that possible.
* **Consequence**: the predicted effect on the gene.

Tandem repeats and methylation findings are placed in a tier by their classification rather than by these factors.

To see the list by classification instead, change **Sort** to **Classification**. This gives a flat list ordered from most to least severe, without tiers or rank numbers.

### Reading a finding

Calls are grouped into one row per variant type and gene. For example, an SNV and an SV in the same gene appear as two rows, and several SVs in one gene appear as one row with an **N calls** pill.

Each row shows:

* its rank and a variant-type badge (SNV, SV, CNV, TR, METH);
* the automated classification chip;
* the gene and locus;
* a one-line description, for example the HGVS change and consequence for a small variant, or the size, type, and genes for an SV;
* a **Network** badge with the HiFi Solves frequency, when available;
* a strength meter.

Coloured **evidence chips** under each row show which factors raised or lowered its rank. The legend at the foot of the card reads **Evidence: raises rank / supports / lowers rank**. Hover a chip for detail.

### Reviewing findings

1. Filter by variant type, if you want to, using the type dropdown (**All types** by default).
2. Click **Show evidence** on a finding to open its evidence ledger. The top-ranked finding is open by default. The ledger explains why the finding ranks where it does, one row per evidence factor.
3. Click **Open variant**, or click the row, to open the variant drawer. Use its previous/next controls to step through the findings in order without leaving the Summary tab.
4. [Classify](/products/workbench/variant-interpreter/classifying-variants.md) the variant. To report it, click **Add to report** on the finding's row. The button changes to **In report**; click it again to remove the variant from the report. See [Filling in a report](/products/workbench/variant-interpreter/generating-a-report/filling-in-a-report.md#adding-variants-to-the-report).

To look beyond the priority findings, click **Search all variants →** to open the [Search tab](/products/workbench/variant-interpreter/reviewing-variants/search.md).

### Calls set aside

Some calls meet the classification criteria but are unlikely to be causal. These are moved into a collapsed panel below the ranked list instead of being ranked. The panel summarizes how many were set aside and why:

* **common in cohort**: frequent in control populations;
* **repeat or segmental-duplication region**: in a region where calls are less reliable;
* **no constrained gene**: no overlapping gene is constrained or has an established disease association.

Click **Show all** to review them. Nothing is removed. Small variants, tandem repeats, and methylation findings are never set aside, and neither is any call you have classified as Pathogenic or Likely Pathogenic.

{% hint style="warning" %}
If part of the case couldn't be read or scored, a warning appears above the list stating that the findings may be incomplete and why. Refresh the page to try again before relying on the list.
{% endhint %}

## Haplotype phase

The **Haplotype phase** card answers the cis-or-trans question for every gene where the priority findings include two or more small variants. It shows how many reported pairs have a phase verdict, and lists the pairs that can and can't be called in trans. See [Haplotype phase (cis/trans)](/products/workbench/variant-interpreter/reviewing-variants/haplotype-phase.md).

## Run quality

The **Run quality** strip summarizes how well the sample was sequenced and phased:

| Tile                     | What it shows                                                                                    |
| ------------------------ | ------------------------------------------------------------------------------------------------ |
| **Mean coverage**        | Average read depth across the genome.                                                            |
| **Het calls phased**     | The percentage of heterozygous calls assigned to a haplotype, with the phase block NG50 beneath. |
| **Clinical loci unread** | How many clinically important loci had no usable coverage (dropouts).                            |

For a family, coverage and phasing are averaged across samples, and **Clinical loci unread** shows the worst sample.

A tile shows **Unavailable** when the analysis didn't produce the quality file it needs. Hover it to see why.

## Technical detail

The **Per-sample QC** table breaks the run quality measures down by sample: **Sample**, **Mean coverage**, **Het calls phased**, **Phase block NG50**, and **Clinical loci unread**. It is hidden when no quality data is available.

## Where to go next

To review a variant type in full, switch to its tab. See [Reviewing variants](/products/workbench/variant-interpreter/reviewing-variants.md). The sample drawer is available here and on every other screen for sample context.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.omics.ai/products/workbench/variant-interpreter/getting-started/the-summary-tab.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.