THPdb: Database of FDA-approved peptide and protein therapeutics

Abstract

THPdb (http://crdd.osdd.net/raghava/thpdb/) is a manually curated repository of Food and Drug Administration (FDA) approved therapeutic peptides and proteins. The information in THPdb has been compiled from 985 research publications, 70 patents and other resources like DrugBank. The current version of the database holds a total of 852 entries, providing comprehensive information on 239 US-FDA approved therapeutic peptides and proteins and their 380 drug variants. The information on each peptide and protein includes their sequences, chemical properties, composition, disease area, mode of activity, physical appearance, category or pharmacological class, pharmacodynamics, route of administration, toxicity, target of activity, etc. In addition, we have annotated the structure of most of the protein and peptides. A number of user-friendly tools have been integrated to facilitate easy browsing and data analysis. To assist scientific community, a web interface and mobile App have also been developed.

Designing Anticancer Peptides by Constructive Machine Learning

“…None of the generated peptides were part of the fine-tuning set, contained in the CancerPPD database, [11] or identicalt oa nF DA-approved/investigational therapeutic peptide. [12] This result corroborates the ability of the machine learning model to generate new amino acid se- quences. Importantly,t he computer-designed sequences inherited properties of the training and fine-tuning sets.…”

“…None of the generated peptides were part of the fine-tuning set, contained in the CancerPPD database, [11] or identicalt oa nF DA-approved/investigational therapeutic peptide. [12] This result corroborates the ability of the machine learning model to generate new amino acid se-Figure 1. Generic scheme of constructivemachine learning for peptide design.T he training was performedi ntwo steps, first with computationally generated helical amphipathic peptide sequences, then with 26 known anticancer peptides (ACPs) for model fine-tuning(transferl earning).…”

“…Of these sequences, 98 % were unique, and only one was identical to a sequence from the training set. None of the generated peptides were part of the fine‐tuning set, contained in the CancerPPD database, or identical to an FDA‐approved/investigational therapeutic peptide . This result corroborates the ability of the machine learning model to generate new amino acid sequences.…”

Jejunal wall delivery of insulin via an ingestible capsule in anesthetized swine—A pharmacokinetic and pharmacodynamic study

“…Since the approval of RHI by the US Food and Drug Administration, nearly 40 years ago, therapeutic proteins and peptides have proliferated to become major players in the management of a variety of common and rare diseases. 1,4,5 One important limitation, however, is the need for many, if not most protein drugs, to be administered parenterally to (a) obviate their destruction by a passage through the acidic gastric environment, (b) promote their access to the bloodstream in high enough concentrations, and (c) preserve their therapeutic activity, factors which all represent an enormous challenge. 1 The relationship between the obligation to self-administer repetitive subcutaneous or intramuscular injections and therapeutic non-compliance has been well documented by previous studies.…”

Minimotifs dysfunction is pervasive in neurodegenerative disorders

“…The Therapeutics Peptide Database has no FDA‐approved drugs for treating NDs [264]. Despite this lack of success, the investigation of minimotif‐directed therapeutics has been, and remains of high interest and potential.…”