BEGIN:VCALENDAR VERSION:2.0 PRODID:icalendar-ruby CALSCALE:GREGORIAN BEGIN:VEVENT DTSTAMP:20261017T121851Z UID:e9c1ddc8-7f17-40ec-8787-3016ea56d071 DTSTART:20261211T090000Z DTEND:20261211T170000Z DESCRIPTION:The course is oversubscribed and has therefore been closed.\n\n ## Overview\n\nFoundation models trained on biological sequences\, often c alled biological language models\, are changing the way we analyse protein s\, DNA and RNA. Built on the same architectures as Large Language Models (LLMs)\, models such as ESM-2 or ProtT5 learn rich representations of sequ ences that can be reused for a wide range of tasks\, from functional annot ation to variant effect and structure prediction.\n\nThis one-day course o ffers a practical introduction to this rapidly evolving field. After a sho rt overview of the main model architectures and of the models available fo r different molecular data types\, participants will work hands-on in Pyth on notebooks to load pre-trained sequence language models with Hugging Fac e Transformers\, extract and explore sequence embeddings\, and fine-tune a model on a downstream prediction task. Short lectures alternate with prac tical exercises run on GPU-enabled notebooks.\n\nBy the end of the course\ , participants will have an overview of the current landscape of biologica l language models and will be able to judge when and how to apply them to their own research questions.\n\n\n## Audience\n\nThis course is designed for PhD students\, postdoctoral and other researchers in the life sciences \, from both academia and industry\, who already have practical experience with deep learning and would like to use pre-trained biological language models in their own analyses.\n\n## Learning outcomes\n\nAt the end of the course\, the participants should be able to:\n\n- **Compare** encoder (e. g. BERT)\, decoder (e.g. GPT) and long-context (e.g. Hyena) architectures\ , and explain which molecular data types (proteins\, DNA\, RNA) and tasks each is suited to.\n- **Apply** pre-trained biological sequence language m odels to generate embeddings from sequences using the Hugging Face Transfo rmers library.\n- **Analyse** embeddings using visualisation\, clustering and classification to explore relationships between sequences.\n- **Adapt* * a pre-trained sequence language model to a downstream prediction task th rough fine-tuning.\n- **Evaluate** whether a biological language model is appropriate for a given research question\, considering data type\, model size\, computational cost and known limitations.\n\n\n## Prerequisites\n\n ### Knowledge / competencies\n\n- A good fluency with the Python programmi ng language\, including working knowledge of common data analysis librarie s such as numpy\, pandas and matplotlib.\n- Prior knowledge of core deep l earning concepts (neural networks\, loss functions\, training loops\, back propagation) and practical experience of training models with PyTorch are required.\n- Familiarity with the Transformer architecture and the attenti on mechanism is recommended.\n- Familiarity with protein sequence data and resources (e.g. FASTA format\, UniProt) is recommended.\n\n\nTo get the m ost out of this course\, you should meet the learning outcomes of [Diving into Deep Learning - Theory and Applications with PyTorch](https://www.sib .swiss/training/course/DEEPP). \nIf you are new to Python or machine learn ing\, we recommend first following [First Steps with Python in Life Scienc es](https://www.sib.swiss/training/course/FSWPY)\, [Data Analysis and Repr esentation in Python](https://www.sib.swiss/training/course/DARPY) and [In troduction to Machine Learning with Python](https://www.sib.swiss/training /course/INMLP).\n\n### Technical\n\nYou will need access to a computer wit h a stable internet connection.\n\nThe practicals run on GPU-enabled Pytho n notebooks in the cloud. You will receive access instructions before the course\; no local installation is required.\n\n\n## Application\n\nThe cou rse is oversubscribed and has therefore been closed.\n\nThe registration f ees for academics are **100 CHF** and **500 CHF** for for-profit companies .\n\nWhile participants are registered on a first come\, first served basi s\, exceptions may be made to ensure diversity and equity\, which may incr ease the time before your registration is confirmed.\n\nApplications will close on **27/11/2026** or as soon as the places will be filled up. Cancel lation after **27/11/2026** will not be reimbursed. Please note that parti cipation in SIB courses is subject to our [general conditions](https://www .sib.swiss/legal-documents).\n\nYou will be informed by email of your regi stration confirmation. Upon reception of the confirmation email\, particip ants will be asked to confirm attendance by paying the fees within **5 wor king days**.\n\n\n## Venue and Time\n\nThis course will be streamed.\n\nTh e course will start at 9:00 CET and end around 17:00 CET.\n\nPrecise infor mation will be provided to the registered participants in due time.\n\n## Additional information\n\nCoordination: Diana Marek\, SIB Training group.\ n\nA **Certificate of Attendance** will be sent provided you were present at the course\, whereas a **Certificate of Achievement** recommending 0.25 ECTS will be sent provided you passed the exam.\n\nYou are welcome to reg ister to the SIB courses mailing list to be informed of all future courses and workshops\, as well as all important deadlines using the form [here]( https://lists.sib.swiss/mailman/listinfo/courses).\n\nSIB abides by the [E LIXIR Code of Conduct](https://elixir-europe.org/events/code-of-conduct). Participants of SIB courses are also required to abide by the same code.\n \nFor more information\, please contact [training@sib.swiss](mailto:traini ng@sib.swiss). SUMMARY:Biological Large Language Models for Bioinformatics URL;VALUE=URI:https://www.sib.swiss/training/course/20261211_BLLMS END:VEVENT END:VCALENDAR