OpenProteinAI · GiorgiaBorgmann · Sep 8, 2025 · Aug 27, 2025 · Aug 27, 2025 · Aug 29, 2025
diff --git a/source/_static/structure-prediction/minifold.png b/source/_static/structure-prediction/minifold.png
diff --git a/source/_static/tools/poet/context-1.png b/source/_static/tools/poet/context-1.png
diff --git a/source/_static/tools/poet/context-2.png b/source/_static/tools/poet/context-2.png
diff --git a/source/_static/tools/poet/context-3.png b/source/_static/tools/poet/context-3.png
diff --git a/source/_static/tools/poet/query-1.png b/source/_static/tools/poet/query-1.png
diff --git a/source/_static/tools/poet/query-2.png b/source/_static/tools/poet/query-2.png
diff --git a/source/web-app/poet/prompts.rst b/source/web-app/poet/prompts.rst
@@ -4,15 +4,162 @@ Prompt and prompt sampling methods
 What is a prompt?
 -----------------
 
-A prompt is an input that instructs a Generative AI model to generate the desired response. PoET uses a prompt made up of a set of related sequences. These sequences may be homologs, family members, or some other grouping that represents your protein of interest.
+A prompt is an input that directs a generative AI model to produce the desired protein sequences. For PoET-2, a prompt can include sequences and/or structures that define the target protein subspace. In contrast, PoET-1 uses a prompt composed of a set of related sequences. These sequences can be homologs, family members, or other groupings that capture the characteristics of the protein of interest.
 
-The prompt can be:
+A PoET-2 prompt is made up of two components, either of which can be included or omitted depending on your use case.
 
-- uploaded directly as a ``.fasta`` or ``.csv`` file
-- a multiple sequence alignment (MSA) uploaded by the user, or
-- created by the PoET model doing a homology search based on a seed sequence provided by the user.
+- Context: Sequences and/or structures that guide PoET-2’s output distribution and enable in-context learning.
+- Query: Specific constraints, such as sequence length, motifs, or active site residues, allowing precise control over the generated sequence.
 
-Uploading a Prompt
+**Note:** PoET-1 is still available for some use cases and legacy workflows, but we recommend PoET-2 for most scenarios.
+
+Creating a query
+-----------------
+A query allows you to specify precise constraints for PoET-2 to follow during sequence generation. 
+
+Query components
+~~~~~~~~~~~~~~~~
+
+- **Reference sequence:** Baseline sequence for comparison and edits.  
+- **Query sequence:** User-defined sequence for masking sites (valid amino acids and gap tokens only).  
+- **Structure:** Defines structural positions to mask (`X` for masked, `S` for unmasked).
+
+The query enables targeted generation tasks such as sequence in-filling, inverse folding, or motif scaffolding. Only **one** sequence or structure can be entered per query.
+
+Uploading a query
+~~~~~~~~~~~~~~~~~
+
+You can enter into the sequence editor or upload a query in the following formats:
+
+- **Sequence-only files (`.fasta` or `.csv`):**  
+  Uploading a sequence file displays both the reference sequence and the query sequence. You can indicate masked positions in the query sequence using the `X` token. The reference sequence serves as a guide to identify positions for masking and to compare edits against the original sequence.
+
+- **Structure files (`.pdb`):**  
+  Uploading a structure file allows you to indicate masked positions in the structure track:  
+  - `X` — mask this position  
+  - `S` — keep the structure corresponding to the amino acid
+
+You also have the option to skip entering a query by toggling the disable query switch.
+
+.. image:: ../../_static/tools/poet/query-1.png
+  :alt: Uploading query
+
+Sequence Editor Tools
+~~~~~~~~~~~~~~~~~~~~~
+
+The sequence editor provides buttons at the top for efficient query editing:
+
+- **Undo / Redo** — revert or restore changes  
+- **Mask Sequence** — mask all positions using the `X` token  
+- **Unmask Sequence** — restore all positions to match the reference sequence  
+- **Mask Selected Residues** — mask only the highlighted positions  
+- **Mask Unselected Residues** — mask all positions except the highlighted ones
+
+Additional keyboard shortcuts include:
+
+- Copy and paste sequences  (Ctrl + C / V)
+- Replace highlighted positions with a character (e.g., highlight positions 1–50 and press `X` to mask that region)
+
+These tools allow precise control over the query, enabling you to define exactly which residues or structural positions should guide PoET-2’s generation.
+
+.. image:: ../../_static/tools/poet/query-2.png
+  :alt: Sequence editor tools
+
+Creating a Context
+~~~~~~~~~~~~~~~~~~~
+
+Users can either upload a custom context or build one from a Multiple Sequence Alignment (MSA).
+
+Custom Context
+^^^^^^^^^^^^^^
+- Users can upload `.fasta`, `.csv`, or `.pdb` files.
+- Multiple files can be uploaded to a single prompt/context.
+
+.. image:: ../../_static/tools/poet/context-1.png
+  :alt: Context empty state
+
+.. image:: ../../_static/tools/poet/context-2.png
+  :alt: Context populated
+
+Build from MSA
+^^^^^^^^^^^^^^^
+There are two ways to create a context from an MSA:
+
+1. **Upload an existing MSA file:** Users can upload an MSA file directly.  
+2. **Run a homology search using a seed sequence:** Users input a single seed sequence, and PoET builds an MSA by searching for homologs.
+
+.. image:: ../../_static/tools/poet/context-3.png
+  :alt: Context build from MSA
+
+Uploading and Saving a Sequence-Only Prompt
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Without a Project
+^^^^^^^^^^^^^^^^^
+On the **Projects** page, select a PoET tool from the navigation bar. Under **Prompt Definition**, click **Select a file** and choose a `.fasta` or `.csv` file. Ensure **Prompt** is selected before uploading.
+
+.. image:: ../../_static/tools/poet/prompt-1.png
+  :alt: Uploading prompt without a project
+
+Within a Project
+^^^^^^^^^^^^^^^^
+Prompts can be uploaded via:
+
+- **Project Page:** Click **Upload**, select **Prompt**, and upload your `.fasta` or `.csv` file.
+
+.. image:: ../../_static/tools/poet/prompt-2.png
+  :alt: Uploading prompt from project
+
+- **Left Sidebar:** Click the **Upload** button under the **Prompt** section and select your file.
+
+.. image:: ../../_static/tools/poet/prompt-3.png
+  :alt: Uploading prompt within a project
+
+.. image:: ../../_static/tools/poet/prompt-4.png
+  :alt: Uploaded prompt preview
+
+- **From a MSA:** On an existing MSA page, click **Create Prompt**.
+
+.. image:: ../../_static/tools/poet/prompt-5.png
+  :alt: Create prompt from MSA page
+
+
+What is a Multiple Sequence Alignment?
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Multiple sequence alignment (MSA) is a technique for biological sequence analysis. It consists of a sequence alignment of three or more biological sequences that usually have an evolutionary relationship.
+
+Why is MSA Useful?
+^^^^^^^^^^^^^^^^^^
+
+The resulting MSA can be used to infer sequence homology and conduct phylogenetic analysis to assess the sequences’ shared evolutionary origins. Biologically sound and accurate alignments show homology and relationships, allowing for new member identification and the comparison of similar sequences. Accuracy is vital because subsequent analyses depend on the MSA results.
+
+When building a prompt from an MSA, include sequences you want to optimize. The model learns the patterns of the proteins and predicts sequences that best fit that list. Since the model views proteins in their entirety, you cannot optimize for a specific property or activity.
+
+
+Creating a Prompt Using a MSA
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+Without a Project
+""""""""""""""""""
+Navigate to any PoET tool under **Prompt Definition**. You can either input the MSA directly or upload an existing `.fa`, `.fasta`, or `.csv` file.
+
+.. image:: ../../_static/tools/poet/prompt-6.png
+  :alt: Uploading MSA without a project
+
+Within a Project
+"""""""""""""""""
+MSAs can be uploaded via:
+
+- **Project Page:** Click **Upload**, select **MSA**, and input or upload a `.fa`, `.fasta`, or `.csv` file.
+
+.. image:: ../../_static/tools/poet/prompt-7.png
+  :alt: Uploading MSA on project page
+
+![Uploading MSA popup on pr]()
+
+
+Uploading and saving a sequence only-prompt
 -----------------
 
 Without a Project
@@ -42,6 +189,7 @@ Prompts can be uploaded via:
 
 .. image:: ../../_static/tools/poet/prompt-5.png
   :alt: create prompt from MSA page
+
 
 What is a Multiple Sequence Alignment?
 --------------------------------------

diff --git a/source/web-app/structure-prediction/using-structure-prediction.rst b/source/web-app/structure-prediction/using-structure-prediction.rst
@@ -19,6 +19,7 @@ We recommend using:
 - AlphaFold2 for predictions where accuracy is more important than speed. AlphaFold2 creates and samples an MSA in order to perform structure predictions, which increases accuracy but is slower than ESMFold. 
 - Boltz-1 focuses on high-accuracy modeling of biomolecular structures — including proteins, DNA, and RNA — and produces static 3D models of molecular complexes with structural accuracy comparable to AlphaFold3.
 - Boltz-2 is the recommended model for proteins, RNA, DNA and ligands. It expands from Boltz-1 from static complexes to dynamic structural ensembles. This means Boltz‑2 can model how biomolecules move and interact over time.
+- MiniFold is a fast single-sequence structure prediction model built on ESM-2, delivering accuracy comparable to ESMFold while reducing inference time by 10–20×. It is designed for rapid prediction of large numbers of protein structures and currently supports single-chain proteins
 
 Accessing the Structure Prediction tool
 ---------------------------------------
@@ -82,6 +83,18 @@ The **Advanced Options** section contains several parameters:
 
 .. image:: ../../_static/structure-prediction/boltz.png
    :alt: Boltz-1 and Boltz-2
+
+Using MiniFold
+-------------
+
+If you select **MiniFold**, the **Advanced Options** let you set the **Number of recycles**, which controls how many times the network refines a structure by feeding the output of one cycle into the next. This parameter defaults to **auto** and accepts integers from **1 to 48**.  
+
+**Additional notes:**  
+- Supports only single-chain proteins  
+- Maximum sequence length: **2048**  
+
+.. image:: ../../_static/structure-prediction/minifold.png
+   :alt: MiniFold
 
 Visualizing your sequence
 --------------------------