Prediction of delayed retention of antibodies in hydrophobic interaction chromatography from sequence using machine learning

Jain, Tushar; Boland, Todd; Lilov, Asparouh; Burnina, Irina; Brown, Michael E.; Xu, Yingda; Vásquez, Maximiliano

doi:10.1093/bioinformatics/btx519

Cited by 51 publications

(48 citation statements)

References 87 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As previously stated, a higher AggScore value relates to a higher tendency to form amyloid fibrils. Since the AggScore method has not been optimized for a particular cutoff to distinguish between amyloidogenic and nonamyloidogenic peptides, we evaluated the performance of the score using the conventional receiver operating characteristic curve (ROC) analysis, as demonstrated in similar studies . For evaluating the performance, the peptides were first sorted in increasing order with respect to their AggScore values.…”

Section: Resultsmentioning

confidence: 99%

AggScore: Prediction of aggregation‐prone regions in proteins based on the distribution of surface patches

et al. 2018

View full text Add to dashboard Cite

Protein aggregation is a phenomenon that has attracted considerable attention within the pharmaceutical industry from both a developability standpoint (to ensure stability of protein formulations) and from a research perspective for neurodegenerative diseases. Experimental identification of aggregation behavior in proteins can be expensive; and hence, the development of accurate computational approaches is crucial. The existing methods for predicting protein aggregation rely mostly on the primary sequence and are typically trained on amyloid-like proteins. However, the training bias toward beta amyloid peptides may worsen prediction accuracy of such models when applied to larger protein systems. Here, we present a novel algorithm to identify aggregation-prone regions in proteins termed "AggScore" that is based entirely on three-dimensional structure input. The method uses the distribution of hydrophobic and electrostatic patches on the surface of the protein, factoring in the intensity and relative orientation of the respective surface patches into an aggregation propensity function that has been trained on a benchmark set of 31 adnectin proteins. AggScore can accurately identify aggregation-prone regions in several well-studied proteins and also reliably predict changes in aggregation behavior upon residue mutation. The method is agnostic to an amyloid-specific aggregation context and thus may be applied to globular proteins, small peptides and antibodies.

show abstract

Section: Resultsmentioning

confidence: 99%

AggScore: Prediction of aggregation‐prone regions in proteins based on the distribution of surface patches

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Hydrophobicity. Hydrophobicity in the CDR regions has been repeatedly linked to aggregation propensity in mAbs (2,(6)(7)(8). Using our homology models, we estimated the effective hydrophobicity of each residue by considering not only its degree of apolarity, but also whether or not it is solvent-exposed (side chain relative ASA rel > 7.5% (21,22)).…”

Section: R a F Tmentioning

confidence: 99%

“…While some cases of poor developability are subtle in origin, others are less ambiguous. High levels of hydropho-bicity, particularly in the highly variable complementaritydetermining regions (CDRs), have repeatedly been implicated in aggregation, viscosity and polyspecificity (2)(3)(4)(5)(6)(7)(8). Asymmetry in the net charge of the heavy and light chain variable domains is also correlated with self-association and viscosity at high concentrations (4,9).…”

Section: Introductionmentioning

confidence: 99%

“…A primary focus in recent years has been in designing software that can better predict aggregation proclivity. Many algorithms designed for this purpose use only the antibody sequence (4,7,8), although some suggest an analogous equation to use if a structure is available (4). One purely structure-based method is the structural aggregation propensity (SAP) metric (5), later included in the Developability Index (6).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The Therapeutic Antibody Profiler (TAP): Five Computational Developability Guidelines

Raybould

Marks

Krawczyk

et al. 2018

Preprint

View full text Add to dashboard Cite

Therapeutic monoclonal antibodies (mAbs) must not only bind to their target but must also be free from 'developability issues', such as poor stability or high levels of aggregation. While small molecule drug discovery benefits from Lipinski's rule of five to guide the selection of molecules with appropriate biophysical properties, there is currently no in silico analog for antibody design. Here, we model the variable domain structures of a large set of post-Phase I clinical-stage antibody therapeutics (CSTs), and calculate an array of metrics to estimate their typical properties. In each case, we contextualize the CST distribution against a snapshot of the human antibody gene repertoire. We describe guideline values for five metrics thought to be implicated in poor developability: the total length of the Complementarity-Determining Regions (CDRs), the extent and magnitude of surface hydrophobicity, positive charge and negative charge in the CDRs, and asymmetry in the net heavy and light chain surface charges. The guideline cut-offs for each property were derived from the values seen in CSTs, and a flagging system is proposed to identify nonconforming candidates. On two mAb drug discovery sets, we were able to selectively highlight sequences with developability issues. We make available the Therapeutic Antibody Profiler (TAP), an open-source computational tool that builds downloadable homology models of variable domain sequences, tests them against our five developability guidelines, and reports potential sequence liabilities and canonical forms. TAP is freely available at http://opig.stats.ox.ac.uk/webapps/sabdab-sabpred/TAP.php therapeutic monoclonal antibodies | developability guidelines | immunoglobulin gene sequencing | surface hydrophobicity | surface charge Correspondence: deane@stats.ox.ac.uk Raybould et al. | bioRχiv | June 29, 2018 | 1-7

show abstract

“…(2017a) dataset, reports have appeared in the literature using the dataset. For example, predictive models of HIC performance using QSPR models (Jetha et al, 2018) and a combined sequence and structure approach (Jain et al, 2017b). CDR properties of the Jain et al (2017a) dataset have also been implicated in identifying antibodies with developmental issues (Raybould et al, 2019), and the dataset has also been used to benchmark aggregation prediction algorithms (Sankar et al, 2018).…”

Section: Introductionmentioning

confidence: 99%

Charge and hydrophobicity are key features in sequence-trained machine learning models for predicting the biophysical properties of clinical-stage antibodies

Hebditch

Warwicker

2019

Preprint

View full text Add to dashboard Cite

Improved understanding of properties that mediate protein solubility and resistance to aggregation are important for developing biopharmaceuticals, and more generally in biotechnology and synthetic biology. Recent acquisition of large datasets for antibody biophysical properties enables the search for predictive models. In this report, machine learning methods are used to derive models for 12 biophysical properties. A physicochemical perspective is maintained in analysing the models, leading to the observation that models cluster largely according to charge (cross-interaction measurements) and hydrophobicity (self-interaction methods). These two properties also overlap in some cases, for example in a new interpretation of variation in hydrophobic interaction chromatography. Since the models are developed from differences of antibody variable loops, the next stage is to extend models to more diverse protein sets. Availability:The web application for the sequence based algorithms are available on the proteinsol webserver, at https://protein-sol.manchester.ac.uk/abpred, with models and virtualisation software available at https://protein-sol.manchester.ac.uk/software.

show abstract

Prediction of delayed retention of antibodies in hydrophobic interaction chromatography from sequence using machine learning

Abstract: Supplementary data are available at Bioinformatics online.

Cited by 51 publications

References 87 publications

AggScore: Prediction of aggregation‐prone regions in proteins based on the distribution of surface patches

AggScore: Prediction of aggregation‐prone regions in proteins based on the distribution of surface patches

The Therapeutic Antibody Profiler (TAP): Five Computational Developability Guidelines

Charge and hydrophobicity are key features in sequence-trained machine learning models for predicting the biophysical properties of clinical-stage antibodies

Contact Info

Product

Resources

About