![[feed]](/style/images/feed-icon-14x14.png){kind=icon}
Tools
Tools
| Item Type: | Dataset |
|---|---|
| Title: | Malva polyA query probes and LLM evaluation |
| Creators: |
León-Periñán, Daniel  ORCID: https://orcid.org/0000-0002-7970-0362, Karaiskos, Nikos ORCID: https://orcid.org/0000-0001-7771-3947 and Rajewsky, Nikolaus ORCID: https://orcid.org/0000-0002-4785-4332
|
| Abstract: | Data deposits for "Malva: ultrafast, large-scale sequence discovery in single cells" This repository contains supplementary data files referenced in the manuscript. Contents: 1. polyA_probes.fa Polyadenylation site detection probes (Sup. Fig. 10) 2. llm_evaluation_queries.json Natural language query interface evaluation dataset (Sup. Fig. 12) 1. Polyadenylation site detection probes for Malva This file contains nucleotide probes designed for detecting alternative polyadenylation events using Malva Index. Probes were constructed from polyASite v3.0 (Moon et al., Nucleic Acids Res. 2025) annotations. Probes with extreme GC content (<30% or >70%) were excluded. Probe headers encode the polyASite identifier, genomic coordinates, probe type (polyA or control), and the associated gene symbol. File format: FASTA (.fa) Reference genome: GRCh38 2. Evaluation dataset for the Malva natural language query interface This file contains 10,000 synthetic test cases used to evaluate the accuracy of the natural language query translation engine in the Malva platform. Test cases were generated using a local instance of gpt-oss:120b, ensuring independence from the production model (Llama 3.1 8B). Each entry contains: - query: the natural language input string - expected_type: the correct query category (one of: simple_gene, gene_with_filters, marker_genes, sequence_search, database_lookup, pathway_query, unsupported, over_inference_trap) - expected_genes: list of gene symbols that should be extracted - expected_filters: dictionary of metadata filters explicitly stated in the query (e.g., tissue, disease, cell_type) - expected_sequence: nucleotide sequence, if applicable The over_inference_trap category (15% of cases) tests whether the model inappropriately adds biological context as filters when not explicitly requested by the user. File format: JSON (.json) |
| Source: | Zenodo |
| Publisher: | CERN |
| Date: | 24 June 2026 |
| Additional Information: | Copyright (C) 2026 The Authors |
| Official Publication: | https://doi.org/10.5281/zenodo.19455194 |
Repository Staff Only: item control page
