Predicting protein structures has become much more accessible thanks to tools like AlphaFold-2. However, setting up a structure prediction pipeline and visualizing results can still demand considerable effort—especially for those new to computational biology or looking to streamline their workflow.
If you’ve ever asked yourself: “Can I just upload a FASTA file and get a reliable 3D structure?” — the answer is yes. With the Biomolecular Structure Prediction extension in SAMSON, you can bring AlphaFold-2 predictions directly into your molecular modeling projects in just a few steps. Here’s how.
Why this matters
AlphaFold-2 has revolutionized how researchers approach protein structure prediction. But the outputs are often buried in web interfaces or require users to build complex pipelines. SAMSON simplifies this process by integrating AlphaFold-2 seamlessly into a graphical, interactive environment. This can be particularly helpful for molecular modelers, students, and educators who value both predictive power and visual interpretability.
Preparing a prediction
To get started in SAMSON:
- Open Home > Predict from the top menu.
- Select the AlphaFold-2 tab from the list of prediction services.
- Upload your FASTA file(s). You can predict single or multiple sequences.
- Choose whether you’re working with monomers or multimers, and select the appropriate multiple sequence alignment (MSA) database.
- Click Start prediction.
Your job will then run on cloud servers, using high-performance computing instances that you can select based on the resources available. The servers include options like those powered by NVIDIA A100 GPUs for increased speed and accuracy.
Visualizing the results
Once the prediction is complete, open Interface > Cloud jobs in SAMSON or visit your SAMSON Connect > Account > Jobs page to retrieve the results.
Predicted structures are automatically color-coded based on pLDDT scores, which indicate the confidence of the prediction at each residue. This visual cue helps quickly assess which parts of the structure are likely to be reliable models versus regions with higher uncertainty.
A note on computing credits
Running AlphaFold-2 predictions in the cloud does require computing credits. Users can request free credits for evaluation purposes or purchase additional credits as needed. All communications are encrypted, and your uploaded data remains protected during transmission and processing.
When to cite
If you’re planning to publish your results, don’t forget that any findings based on AlphaFold-2 predictions should cite the relevant literature, including the original AlphaFold-2 paper from Nature (2021) and the AlphaFold-Multimer publication if applicable.
Simplifying your workflow
The integration of AlphaFold-2 in SAMSON significantly reduces the barrier to entry for protein structure prediction. Instead of juggling multiple tools, you can go from a FASTA file to an interactive 3D model with just a few clicks—all inside a platform designed for integrative molecular design.
To learn more about using AlphaFold-2 in SAMSON and explore additional features like job sharing and interaction analysis, visit the full documentation at https://documentation.samson-connect.net/tutorials/bsp/bsp/.
SAMSON and all SAMSON Extensions are free for non-commercial use. You can get SAMSON at https://www.samson-connect.net.