深層学習を用いたHLA imputation法の開発とParkinson病と1型糖尿病の原因遺伝子変異解明への応用
深層学習を用いたHLA imputation法の開発とParkinson病と1型糖尿病の原因遺伝子変異解明への応用
A plain-language reading has not been prepared for this paper yet.
THE PARKINSONS LIBRARY
Research made
understandable.
Explore what researchers are learning about Parkinson’s: what may cause it, how it progresses, and what might help stop it.
Publication records to explore
Full article texts acquired
A growing collection across languages and research areas. Many records currently contain only publication details or an abstract.
FIND YOUR STARTING POINT
Four ways into the research.
01 / SEARCH
Explore the papers
Look up a topic, gene or publication. Narrow the collection by year, language or research focus.
Browse the catalog ↗ 02 / UNDERSTANDRead in plain language
Follow the research question, experiment and findings in a short overview or a more detailed story.
Choose a reading ↗ 03 / CONNECTSee the bigger picture
Follow avenues of research through time and discover which topics researchers are studying together.
Open the research map ↗ 04 / COMPARELook closer at results
Explore an early set of extracted findings, including favorable outcomes and results with no clear difference.
Explore the findings ↗START READING
A closer look at the work.
These readings explain acquired full-text papers. AI-authored explanations have not yet been expert-reviewed.
BUILT AROUND THE ORIGINAL RESEARCH
Read clearly.
Follow the evidence.
The library brings publications together to help people understand the research and investigate possible connections. Every explanation keeps its source close by.
A result in cells or animals is different from a benefit shown in people. We keep those distinctions visible, along with negative results and unanswered questions. The collection and its automated classifications are still being developed.
A SHARED BODY OF KNOWLEDGE
Understand the causes.
Find a path forward.
A traceable library for exploring Parkinson's mechanisms,
prevention, and the course of disease.
↗Collect broadly.
Connect carefully.
Connect carefully.
START EXPLORING
A growing collectionTopic tags are unreviewed keyword matches. They indicate where to look, not what has been proven. Obvious WPW heart-rhythm matches are hidden by default.
Research catalog
Newest publication year firstOUTCOMES AND EXPERIMENTS
Explore the findings.
Search what changed, in whom, and under which conditions. A favorable result means improvement on a measured outcome in that experiment; it does not establish a cure or benefit in people.
Early extraction pilotMost papers have not been processed. These records are not expert-reviewed. Matching quotations establish where text came from, not whether an interpretation is correct. Keep negative and uncertain results in view.
RESEARCH, IN EVERYDAY LANGUAGE
Understand a paper.
Open a paper from the catalog, paste its text, or load a PDF. Explore the question, experiment and reported findings with the source beside you.
Explore another example or bring your own paper
A HISTORY OF THE COLLECTED LITERATURE
Avenues of research.
Follow research from earlier publications to recent work. Choose an avenue or a point in time to explore its papers, the teams named in them, and connections to other topics.
These are overlapping topic matches in our collection, not a ranking of treatments or a history of proven discoveries. Collection coverage varies by period and language. A shared topic does not establish a scientific connection.
Recent areas of focus
Research through time
Read left to right. Each cell shows a publication count; darker shading means a larger share or count. Select any cell to open its papers. Papers can appear in several rows.
Explore an avenue
How this map is made and what it cannot show
Avenues use the catalog’s candidate topic tags plus explicit terms in titles and abstracts, including some historical names and non-English terms. Open “Why this paper appears here” to see a match. Rules are incomplete and can match background discussion or a negative result. No model infers a paper’s findings from its title.
The earliest date shown is the earliest matching publication in this library, not the discovery date. Research groups are shown through the authors on individual papers; names have not been combined into laboratories or disambiguated. Shared-topic counts describe publications, not independent experiments, collaboration or causation. Sparse areas may reflect collection gaps rather than neglected scientific opportunities.
YOUR RESEARCH LIBRARY
Welcome.
New in your library
A GROWING APPENDIX
Terms made clear.
Plain-language explanations of biology, experiment design and statistics. Definitions are editorial drafts, with background references.
COVERAGE WITHOUT GUESSWORK
Sources & imports.
Source records can overlap. A language is not an author's country, and a connected source is not an exhaustive collection.
Bring in an authorized database export
Import UTF-8 RIS, CSV, or Web of Science tagged text. Preview the records before adding them to the library.
TRANSPARENT BY DESIGN
Collection history
Queries, source counts, and checkpoints for every collection run.
A bounded sample is not the entire literature. Source counts can overlap and change over time.
PURPOSE & METHOD
Evidence before inference.
This library prioritizes causes, mechanisms of degeneration, prevention, disease modification, and durable restoration. Symptom-management research stays in the collection because useful connections can cross research boundaries.
What is working
Five metadata collectors, multilingual discovery, preserved source responses, identifier-based linking, authorized export imports, permitted PMC article text, PDF/text reading guides, a glossary, source-linked plain-English examples, a durable reading queue and an early searchable findings pilot.
What comes next
Broader licensed access, more full text, trial and dataset links, historical scans, expert review, and broader structured extraction. Study and participant overlap must be assessed before reports can be counted as independent evidence.
What a tag means
Tags are transparent keyword candidates, not an assessment of causality, treatment efficacy, or study quality. A passage saying an approach failed may receive the same topic tag as one reporting success. Open a record to see the words that matched.
Collection boundaries
This release contains bounded API samples. Absence from the catalog means only that material has not been collected here. Some permitted full texts have been acquired; most records still contain metadata or abstracts. Participant-level data have not been acquired. Original abstract text may carry its own reuse restrictions.