Elsevier Author Aid Human Diseases Collection, part 2 (Monogenic Rare Diseases)

Published: 28 May 2024| Version 2 | DOI: 10.17632/8py7y8vvtb.2
Contributors:
,
,
,
,
,

Description

-----> Human Disease Author Aid Collection combines information about rare and common diseases in standardized, easy-to-navigate overview templates and tables. It includes clinical, molecular, and pharmacological data from several Elsevier's and public sources. Tables are planned to be updated quarterly. -----> Author Aid Templates can be a helpful guide for authors, researchers, clinicians, and students, especially those interested in rare diseases, because it highlights updates and findings from several sources on one page. -----> Each disease template overview includes 6 sections: Terminology; Epidemiology/Demographics; Clinical presentation/Diagnosis; Etiology/Pathology (genetics, biomarkers, pathways); Treatment/Follow-Up; Case studies. Each subset of data is linked to a list of publications with relevant citations. -----> Monogenic Rare Diseases (Human Disease Author Aid Collection, Part 2) In the current part, monogenic rare diseases were chosen based on their classification, prevalence, and degree of data availability. By "monogenic" we mean diseases that are caused by one or few known mutations. We considered worldwide point prevalence between <1/10 or 100 thousand (worldwide) as a criterion for filtering most significant rare diseases. Contents A) Curated templates about monogenic rare diseases: • Achondroplasia • Facioscapulohumeral dystrophy • Familial Mediterranean fever • Hereditary breast and ovarian cancer syndrome • Huntington disease • Congenital hypothyroidism • Rett syndrome • Usher syndrome • Wilson’s disease • Wiskott-Aldrich syndrome B) Automatically collected information about additional 21 Rare monogenic diseases. -----> PLEASE download files to read them and open the links!

Files

Steps to reproduce

Automatic excerption followed by manual curation is used to generate the templates. Automatic information excerption is powered by Elsevier Natural Language Processing technology, products, and public sources (ClinicalKey & EMMET, Pathway Studio, Pharmapendium, Elsevier Text Mining, Orphanet, ClinicalTrials, PubMed, and others). Examples of queries and python code to retrieve network information from ETM can be found on GitHub [https://github.com/nesterova-anastasia/authoraid-diseases].

Institutions

Elsevier BV

Departments

Life Sciences Solutions

Categories

Genetics, Health Sciences, Orphan Disorder

Licence