Sold by: Genomic Intelligence
Deployed on AWS
Free Trial
Predicts how strongly a human gene is expressed in a given cell type or tissue, directly from its DNA sequence. Send the gene's sequence, the position of its transcription start site and a free-text description of the cell context; get back the predicted expression level. JSON in, JSON out. For research use.
Overview
Predict gene expression from DNA in the specified biological context.
Genomic Intelligence Gene Expression Prediction estimates how strongly a human gene will be expressed from its DNA sequence in a specific cell type, tissue, or experimental condition. Ask about hepatocytes, erythroblasts, macrophages, a treatment condition, or a more detailed experimental context: the model combines the DNA sequence with your biological description to predict expression.
What you can do. Compare the same gene across cell types and tissues. Rank genes by expected expression in a biological context. Test how a regulatory variant may increase or decrease predicted expression by comparing reference and alternative sequences. Use expression predictions to prioritize candidates before moving into more expensive experimental validation.
How it works. Send the model a DNA sequence, the position of the transcription start site (TSS), and a free-text description of the biological context. For example: "hepatocyte", "erythroblast", or a detailed description of an RNA-seq experiment. You can optionally specify the end of the gene to improve sequence selection. The model returns predicted expression in TPM and log-scaled units, together with the sequence region used for the prediction.
Built for biological workflows. The model combines a DNA language model with a text encoder for biological context and was trained on approximately 100 million samples. This lets you specify the biological setting naturally rather than selecting from a fixed list of predefined tracks.
Run it your way. Use the same request format with a real-time endpoint for interactive analyses or batch transform for large-scale screens. Deployment runs inside your own AWS account, so your genomic data stays within your environment.
Get started quickly. The examples repository includes a ready-to-run notebook and sample requests using APOA1 and HBB in hepatocytes and erythroblasts, making it easy to adapt the workflow to your own genes and contexts.
Performance. Approximately 0.3-0.6 seconds per prediction on ml.g5.xlarge.
Research use only. This product is intended for research use and is not a medical device or intended for diagnostic or clinical decision-making.
Highlights
- Predicts expression of a human gene in the cell type or tissue you describe, directly from DNA sequence.
- Simple JSON interface: DNA sequence, transcription start site and a free-text cell context in; one expression value out.
- Real-time endpoint or batch transform on GPU instances, with a sample notebook and ready-to-run example requests.
Details
Sold by
Categories
Delivery method
Latest version
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
Free trial
Try this product free for 14 days according to the free trial terms set by the vendor.
Pricing is based on actual usage, with charges varying according to how much you consume. Subscriptions have no end date and may be canceled any time.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimension | Description | Cost/host/hour |
|---|---|---|
ml.g5.xlarge Inference (Batch) Recommended | Model inference on the ml.g5.xlarge instance type, batch mode | $6.00 |
ml.g5.xlarge Inference (Real-Time) Recommended | Model inference on the ml.g5.xlarge instance type, real-time mode | $6.00 |
ml.g4dn.xlarge Inference (Batch) | Model inference on the ml.g4dn.xlarge instance type, batch mode | $2.00 |
ml.g4dn.xlarge Inference (Real-Time) | Model inference on the ml.g4dn.xlarge instance type, real-time mode | $2.00 |
Dimensions summary
You pay by the host hour, so charges scale with how long each inference instance runs. Pricing splits across two choices. First, you pick an instance type: the ml.g4dn.xlarge or the ml.g5.xlarge, which differ in their underlying compute hardware. Second, you pick a mode: batch, for processing groups of sequences at once, or real-time, for on-demand single predictions. Each combination of instance type and mode is priced separately, giving you four options. There is no upfront commitment; you are billed for the hours you use.
Top-of-mind questions for buyers
Both meter host hours, so you pay for running time either way. Batch mode processes groups of sequences in one job, suiting scheduled bulk work. Real-time mode keeps the instance ready for on-demand single predictions, suiting interactive or pipeline calls. Pick the mode matching how often you send sequences.
A host hour is one hour of one running inference instance. For real-time mode, the endpoint accrues charges while provisioned, even between predictions, because the host stays ready. Batch mode charges only while a job runs. Stopping or deleting the endpoint ends software charges.
The two options differ in their underlying compute hardware, each priced separately by host hour. Your choice affects prediction speed for large sequences. Since expression tasks accept sequences up to 500,000 base pairs, faster hardware can shorten each job, changing how many host hours you accrue.
docs.genomicintelligence.ai+1
Helpful?
Vendor refund policy
Software charges are billed hourly through AWS Marketplace. Use the 14-day free trial to evaluate the product. If you were charged in error, for example for an endpoint that never became ready or a job that failed because of our container, email contact@genomicintelligence.ai within 30 days of the charge with your AWS account ID and the endpoint or job name. Approved refunds of software charges are issued through AWS Marketplace. AWS infrastructure charges cannot be refunded by us.
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
Amazon SageMaker model
An Amazon SageMaker model package is a pre-trained machine learning model ready to use without additional training. Use the model package to create a model on Amazon SageMaker for real-time inference or batch processing. Amazon SageMaker is a fully managed platform for building, training, and deploying machine learning models at scale.
Deploy the model on Amazon SageMaker AI using the following options:
Version release notes
tes_index is now optional; sending it is recommended. Requests that include tes_index return the same predictions as before.
Additional details
Inputs
- Summary
One JSON object per request, Content-Type application/json. The same body works for a real-time endpoint and for each S3 object in a batch transform job.
- sequence (required): DNA containing the gene of interest, in the gene's orientation, with at least 40,960 bp upstream of the transcription start site.
- tss_index (required): 0-based position of the transcription start site in sequence.
- tes_index (optional): 0-based, exclusive end of the gene in sequence, if the sequence includes it.
- options.description (required): free-text description of the cell type, tissue or experimental context, for example "hepatocyte".
- sequence_name (optional): a label, echoed back.
Example:
{"sequence": "ACGT...", "sequence_name": "APOA1_hepatocyte", "tss_index": 40960, "tes_index": 43160, "options": {"description": "hepatocyte"}}
- Limitations for input type
- Include at least 40,960 bp of genomic sequence upstream of the transcription start site, so the model sees the regulatory context. Maximum payload 6 MB.
- Input MIME type
- application/json
https://github.com/genomicintelligence/sagemaker-marketplace-examples/blob/main/expression/data/input/APOA1_hepatocyte.json
https://github.com/genomicintelligence/sagemaker-marketplace-examples/tree/main/expression/data/input
Support
Vendor support
Support is by email at contact@genomicintelligence.ai .
Include the product name, the model version, the instance type, the request body that produced the problem, and the response or the error you got back. If the sequence is confidential, send the length and the first and last 50 bases rather than the sequence itself; that is usually enough to reproduce.
We answer support email on UK business days. We aim to reply within two business days. There is no contractual response time and no uptime commitment: the model runs in your account, on infrastructure you control, so availability is yours and AWS's rather than ours.
We can help with the request and response format, the meaning of the output, deployment and instance choice, and errors coming from the container. We cannot interpret your biology for you, and we cannot advise on clinical or diagnostic use, which this product is not licensed for.
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
NOTE: This deployment requires a specific SageMaker Inference AMI selection (al2-ami-sagemaker-inference-gpu-3-1). Please use the example notebook provided at https://github.com/NVIDIA/nim-deploy/blob/main/cloud-service-providers/aws/sagemaker/awsmarketplacenotebooks/nim-evo2-40b-v2-1-0awsmarketplace.ipynb for deploying the endpoint.
Evo 2 is a biological foundation model that can interpret and generate DNA sequences across various biological scales: from individual molecules to entire genomes while retaining sensitivity to single-nucleotide changes, enabling zero-shot predictions and complex biological system designs.
Open-source NER model for gene entities in biomedical and clinical text. Trained on BC2GM and optimized for state-of-the-art precision, it enables reliable extraction with fast, easy deployment via Hugging Face Transformers.
Open-source NER model for gene entities in biomedical and clinical text. Trained on BC2GM and optimized for state-of-the-art precision, it enables reliable extraction with fast, easy deployment via Hugging Face Transformers.
Open-source NER model for genetics entities in biomedical and clinical text. Trained on GELLUS and optimized for state-of-the-art precision, it enables reliable extraction with fast, easy deployment via Hugging Face Transformers.
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.