Insertions and deletions (indels) in protein sequences alter the residue spacing along the polypeptide backbone and consequently open up possibilities for tuning protein function in a way that is inaccessible by amino acid substitution alone. We describe an optimization-based computational protein redesign approach centered around predicting beneficial combinations of indels along with substitutions and also obtain putative substrate-docked structures for these protein variants. This modified algorithmic capability would be of interest for enzyme engineering and broadly inform other protein design tasks. We highlight this capability by (1) identifying active variants of a bacterial thioesterase enzyme (‘TesA) with experimental corroboration, (2) recapitulating existing active TEM-1 β-Lactamase sequences of different sizes, and (3) identifying shorter 4-Coumarate:CoA ligases with enhanced in vitro activities toward non-native substrates. A separate PyRosetta-based open-source tool, Indel-Maker (http://www.maranasgroup.com/software.htm), has also been created to construct computational models of user-defined protein variants with specific indels and substitutions.

蛋白质序列中的插入和缺失（indels）改变了多肽骨架上的残基间距，因此开辟了以仅靠氨基酸取代无法达到的方式调整蛋白质功能的可能性。我们描述了一种基于优化的计算蛋白质重新设计方法，其中心是预测插入缺失和替换的有益组合，并获得这些蛋白质变体的假定底物对接结构。这种修改后的算法能力将引起酶工程的兴趣，并广泛地告知其他蛋白质设计任务。我们通过 (1) 通过实验确证鉴定细菌硫酯酶 ('TesA) 的活性变体，(2) 概括现有的不同大小的活性 TEM-1 β-内酰胺酶序列，对非天然底物的体外活性。还创建了一个单独的基于 PyRosetta 的开源工具 Indel-Maker (http://www.maranasgroup.com/software.htm)，以构建具有特定插入和替换的用户定义蛋白质变体的计算模型。