Sequence based drug design

High-throughput antibody design using NGS and bigdata analysis

Why Molcure?

Biomolecule big data and AI algorithm for biomolecule discovery / engineering

  • NGS combination


  • Antibody big-data


  • AI design know-how


  • Consensus frame extraction

    Extraction of consensus sequences from antibody DNA sequences that is pooled in native blood
  • Name
  • Name
  • Translation method

    Translate massive NGS DNA sequence data with avoiding errors such as frame-shifting
  • Preprocessing method

    Merge NGS pair-end sequence data using QC appending algorithm for extracting highly reliable sequences
  • Name
  • Name
  • Affinity prediction based on amplification rate

    Cluster amplification rate:
    Calculate normalized amplification rate from pre&post-panning NGS data to make binding affinity index.
  • Clustering method

    Increase accuracy of mapping sequence on antibody space via clustering based on edit distance. Extract important motifs for binding by internal cluster analysis
  • Name
  • Name
  • Training data

    Predicting non-specific antibody sequence by using machine learning algorithm with FDA approved antibody sequence and Molcure’s own non-specific antibody database
  • Feature extraction for machine learning

    Extract from antibody sequences for machine learning.
  • Name

Machine learning algorithms

Develop classifiers via SVM /Random forest / DNN

  • DNN


  • SVM