Central storage system for proprietary antibodies

Store and analyse sequences and 3D structures

Robust, automated, antibody numbering algorithms

Key tools for analysing frameworks, loops, humanness, atypical residues and more.

Integrate your high value proprietary data with our robust statistical framework.

Driven by proven abYsis technology.

ebisu group

chemogenomix © 2012


structural biology


abYsisProprietory allows users to perform all and more of the public version’s functions behind their own firewall.

abYsis allows users to analyse available antibody sequences and structures as well as to compare their own sequence with statistics generated from the large-scale public information.

Integrated antibody sequence & structure analysis

Central to abYsis is the ability to generate position-specific distributions from statistical information derived from analysis of large scale resources.

This enables you to assess overall ‘humanness’, identify atypical characteristics in individual sequences and differences between the repertoires of different organisms.

Germline identification and structure modeling tools enable further in depth analyses.

abYsis encompasses a comprehensive set of antibody sequences, selected from the total EMBL/Genbank repository, plus processed data from the PDB.

Your own proprietary data can be incorporated too.

Numbering and CDR/framework assignment

Several numbering systems are in use to describe the same structural location in different antibodies. Manual numbering is error prone, but exactness is critical as numbering schemes are referenced in key patents.

Assessment of overall ‘humanness’
Even ‘fully human’ antibodies are not perfect and can elicit anti-antibody responses. abYsis embodies new approaches to assess overall ‘humanness’ and can identify specific epitope-like regions that might confer an anti-antibody effect.

Atypical residues and contributions to ‘humanness’

Based on the ability to generate position-specific distributions, abYsis identifies atypical characteristics in individual sequences.

Identification of germline origin

Identifying the germline origins of your antibodies is important as some germlines express better than others. It also helps to determine which mutations have been introduced during somatic hypermutation maturation. This also assists in the ‘germlinealization’ approach to humanization, by finding the nearest human germline to the mouse donor.

Structural analysis

Back mutation for humanization poses the risk of changing loop conformation. By monitoring the predicted conformation as residues are modified this risk can be decreased. abYsis encompasses:

● Canonical structure prediction based on sequence and knowledge of key residue frequencies

● Structural searches using processed PDB files for model building

A closer look at Canonical loop Prediction

Examples of predictions based on sequence for canonical loop structures. Summary data is same but top figure shows detailed possibilities for CDR-L1 whereas the bottom figure shows equivalent output for CDR L2.

Access the power of abYsis with abYsisProprietary

Store proprietary sequences, build position dependent matrices even more in line with company’s preferred frameworks.
Generate additional web services or use SQL directly on the schema.

new for 2014
smart fast antibody modelling tool providing robust results to both beginners and experts


    Robust numbering,

    framework and

    CDR definitions

    included for



    Kabat, Chothia,




    Wealth of tools for

    examining unusual

    residue positions,



    Angstrom distance

    3D structure




    A total system for

    both genomics,

    biologist and

    structure based


Data storage



    storage for    


    DNA, protein &

    3D data.

    Combine with

    large-scale public 

    data for complete





    based statistics

    embedded in

    abYsis guide your



    proprietary data

    for favoured