FAQ

Channelpedia is maintained and supported by BBP / EPFL, Switzerland.

Our group (Membrane Systems) in BBP is systematically characterizing ion channels in standardized conditions and controlled environment. All electrophysiology data from Kv channels have been made available. The characterization of K2P, Kir and HCN channels are in progress and characterization of KCa, Ca and TRP channels are planned for the coming years.

All registered users can edit the wiki-like sections, just clicking on Edit at the top-right of each sub-section.

Please contact bbp.channelpedia@epfl.ch for details on automated patch-clamping methods.

ChannelAID is an AI expert on ion channel data. It delivers information supported by verifiable sources.

ChannelAID has access to two different databases:

• a fully open-access database with 1 million academic papers, full-text or abstracts, specifically focused on ion channel literature.
• the Channelpedia database, with the papers selected over the years by the researchers in our lab (see PubMed search).

ChannelAID is useful for:

• a quick and accurate answer to your complex/specific question.

• finding relevant papers related to your research.

ChannelAID is a RAG tool. It searches for related paragraphs in the database by using a two step process:

1. keywords filtering: a bm25 algorithm filters the best 400 relevant paragraphs.
2. semantic ranking: the 7 best ranking documents are selected from the 400 paragraphs.

Channelaid uses a generative model (LLM) to generate the final answer. As discussed in this article, answers are highly dependant on the question/prompt quality. To see the best practice for writing your prompt we suggest to follow their recommendations.

Users of ChannelAID may provide input to the Services (“Input”), and receive output from the Services based on the Input (“Output”). Input and Output are collectively “Content.” You are responsible for Content, including ensuring that it does not violate any applicable law or these Terms. You represent and warrant that you have all rights, licenses, and permissions needed to provide Input to our Services.

Artificial intelligence and machine learning are rapidly evolving fields of study. We are constantly working to improve ChannelAID and make it more accurate, reliable, safe, and beneficial. Given the probabilistic nature of machine learning, use of our Services may, in some situations, result in Output that does not accurately reflect real people, places, or facts.

When users use our Services users understand and agree:

• Output may not always be accurate. Users should not rely on Output from our Services as a sole source of truth or factual information, or as a substitute for professional advice.

• Users must evaluate Output for accuracy and appropriateness for their use case, including using human review as appropriate, before using or sharing Output from the Services.

• Users must not use any Output relating to a person for any purpose that could have a legal or material impact on that person, such as making credit, educational, employment, housing, insurance, legal, medical, or other important decisions about them.

• Our Services may provide incomplete, incorrect, or offensive Output that does not represent EPFL’s or Blue Brain Project’s views. If Output references any third party products or services, it doesn’t mean the third party endorses or is affiliated with EPFL or Blue Brain Project.

Channelpedia is a freely-accessible information management framework for comprehensive ion channel information. We encourage all researchers to share their ion channel data through Channelpedia. At present an automatic data upload is not provided, but we can upload your data to Channelpedia: please contact us at bbp.channelpedia@epfl.ch.

Yes, we have an infrastructure suited to share large amount of data. Please contact us at bbp.channelpedia@epfl.ch, indicating which dataset you are interested for downloading.

Yes, please contact us at bbp.channelpedia@epfl.ch.

The cell ID is a unique ID assigned to each cell as soon as it is recorded; it is the main reference to relate every file and every image.

CHO is the main host cell that we have used for generating stable cell lines over-expressing only one type of ion channel, for the whole Kv family. However, we have also used HEK and CV1 cell lines to over-express some of the Kv channels.

Cell line datasheets contain details on the PCR conditions used for the amplification of ion channel genes and the validation of each cell line by RT-PCR.

Yes, all cell lines are freely available for collaborative or specific projects.

Yes, we welcome all proposals for collaborative projects.

For human, experimental data are available for Kv1.1, Kv1.4, Kv1.5, Kv1.6, Kv2.1, Kv3.4, Kv6.4, Kv11.1 Kv12.3 ion channels.

For mouse, experimental data are available for Kv1.1, Kv1.4, Kv1.5, Kv1.6, Kv2.1, Kv3.4, Kv6.4, Kv10.1 and Kv12.3 ion channels.

An NWB file is a file implemented in the Hierarchical Data Format (HDF5), containing data following the Neurodata Without Borders: Neurophysiology format. The Neurodata Without Borders: Neurophysiology project aims to develop a unified, extensible, open-source data format for cellular-based neurophysiology data.

Each NWB file, which is designed to store data in a self-documenting way, contains all the data related to an experiment. The file can be opened with and HDF5 viewer (e.g. HDFView). In the folder acquisition/timeseries there is a subfolder for each applied stimulus, which contains a subfolder for each repetition, where the dataset of the recordings (data) are stored, along with many other metadata, as the seal and the series resistance (r_series) of the patchclamping. Several other metadata are stored in the general/cell_info folder, as the host_cell and the species of the gene.

Raw data from HEKA Patchmaster are converted into NWB files through custom C libraries (available upon request).

You can plot ion channel data directly with HDFview, or you can use our API for more elaborated plots.

We provide API for Matlab and Python. If you prefer another programming language, please contact us at bbp.channelpedia@epfl.ch.

The criteria to define a good recording are based on 6 parameters, whose values have to be within an empirically-determined range:

• Activity index between 0.3 and 1
• Voltage offset between -15 and 35 mV
• Seal larger than 200 MΩ
• Series resistance smaller than 15.5 MΩ
• Slow capacitance smaller than 35 pF
• Maximal current larger than 0.5 nA

Ion channel models will be soon available for Kv channels.

Ion channel visual representations were generated based off of the amino acids (aa) sequences available on Uniprot, under the section "Stucture".

The aa sequence lengths for each individual section (helix, transmembrane, etc.) were downloaded.
These were converted using a 20/67 scale to generate point (pt) lengths.
The pt sections lengths were then drawn using Adobe Illustrator.

A 20/67 scale was chosen (20 aa = 67 pt) as it delivered the best final proportionate image.

The size of the ion channels was estimated based of the predicted structures available on AlphaFold.
The AlphaFold PDB file for each ion channel was downloaded for human, mouse, and rat.
PDB files were then run through a script (AlphaFold_Reader.py) to create 3D surface projections along the z-axis vector onto a x-y plane.
Generated 2D surface projections were measured to give an estimate of ion channel size in (Å^2)
Surface values were collected and outputted as a single excel file then formatted to our website.
Diagram of methodology

As the Alphafold structure is only a predicted structure generated using AI and not a complete resolution of the protein, it is possible that there may be some level of error in the size value.
This must be taken into account when using these values (model building, etc.)