文献、预印本与临床研究查询：参数与补充说明

用法和示例见功能教程。

各功能按作用、API、输入、输出、示例代码和结果说明介绍。运行前按说明准备凭据、本地文件与可选依赖；远程结果随数据库版本与查询时间变化。

按关键词检索 arXiv 论文

dnakit.databases.ar_xiv_search_papers(*, query, limit=None, sort_by=None, sort_order=None, date_from=None, date_to=None, category=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
query	string	必填	Search query for arXiv papers. Unquoted, unprefixed space-separated keywords are ANDed together (every keyword must match) -- adding more keywords NARROWS results, it does not broaden them. To search multiple synonyms/phrases and broaden recall, join them with explicit OR, e.g. '"machine learning" OR "deep learning" OR "neural networks"'. Supports field prefixes (au:, ti:, abs:, cat:, all:) and AND/OR/ANDNOT boolean operators, which are passed through as-is.
limit	integer	None	Number of papers to return. This sets the maximum number of papers retrieved from arXiv.
sort_by	string	None	Sort order for results. Options: 'relevance', 'lastUpdatedDate', 'submittedDate'
sort_order	string	None	Sort direction. Options: 'ascending', 'descending'
date_from	string	None	Filter results from this date (format: YYYY-MM-DD). Uses submittedDate, the date of the paper's FIRST version -- a paper revised/updated after this date but originally submitted earlier is still excluded.
date_to	string	None	Filter results up to this date (format: YYYY-MM-DD). Uses submittedDate, the date of the paper's FIRST version -- a paper revised/updated after this date but originally submitted earlier is still included.
category	string	None	分类或结果类别，按当前接口的枚举/来源定义填写。
verbose	bool	True	运行提示开关。

按 DOI 查询 bioRxiv 预印本元数据

dnakit.databases.bio_rxiv_get_preprint(*, doi, server=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
doi	string	必填	bioRxiv or medRxiv DOI. Can be full DOI (e.g., '10.1101/2023.12.01.569554') or just the numeric part after '10.1101/' (e.g., '2023.12.01.569554'). Find DOIs using EuropePMC_search_articles, web_search, or from paper citations.
server	string	None	Server to query: 'biorxiv' for bioRxiv preprints or 'medrxiv' for medRxiv preprints. Default is 'biorxiv'.
verbose	bool	True	运行提示开关。

按日期范围列出近期预印本

dnakit.databases.bio_rxiv_list_recent_preprints(*, start_date=None, end_date=None, server=None, cursor=None, query=None, limit=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
start_date	string	None	Start date in YYYY-MM-DD format (e.g., '2024-01-01'). Date range must not exceed 60 days.
end_date	string	None	End date in YYYY-MM-DD format (e.g., '2024-01-03'). Date range must not exceed 60 days.
server	string	None	Server: 'biorxiv' for biology preprints, 'medrxiv' for health sciences preprints
cursor	integer	None	Pagination cursor (0 for first 100 results, 100 for next 100, etc.)
query	string	None	检索词或查询表达式；语法由当前数据源定义。
limit	integer	None	当前请求的返回数量上限；不是全库命中总数。
verbose	bool	True	运行提示开关。

按疾病、干预或申办方检索临床研究

dnakit.databases.clinical_trials_search_studies(*, operation=None, query_cond=None, query_intr=None, query_term=None, filter_status=None, filter_phase=None, filter_study_type=None, page_size=None, next_page_token=None, query=None, condition=None, status=None, max_results=None, intervention=None, sponsor=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
operation	string	None	Operation (optional; defaults to 'search' for this tool).
query_cond	string | null	None	Disease or condition to search for (e.g., 'breast cancer', 'type 2 diabetes', 'COVID-19'). Supports Boolean operators (AND, OR, NOT). Multi-word values are matched as an exact phrase (e.g. 'kidney transplantation' is sent as "kidney transplantation"), which is narrower than a loose keyword match. For looser matching, wrap the value in parentheses (e.g. '(kidney transplantation)') to send it verbatim. Values already containing Essie syntax (AND/OR/NOT/AREA[]/quotes) are passed through unchanged. See executed_query in the response for what actually ran.
query_intr	string | null	None	Intervention or drug to search for (e.g., 'pembrolizumab', 'metformin', 'chemotherapy'). Matched against each study's registered intervention names and synonyms (InterventionName/InterventionOtherName), not free text in study summaries, and multi-word values are matched as an EXACT PHRASE -- so 'donor derived cell therapy' matches 0 studies even though 251 trials match the same words loosely. For looser matching, wrap the value in parentheses (e.g. '(donor derived cell therapy)') to send it verbatim, or use ClinicalTrials_search_studies with query_term for free text across all study fields. Prefer the generic/INN name over a brand name. See executed_query and relaxed_match_check in the response.
query_term	string | null	None	Free-text search across all study fields. Use for general keyword search.
filter_status	string | null	None	Filter by recruitment status. Options: 'RECRUITING', 'NOT_YET_RECRUITING', 'ACTIVE_NOT_RECRUITING', 'COMPLETED', 'WITHDRAWN', 'SUSPENDED', 'TERMINATED', 'ENROLLING_BY_INVITATION'. Comma-separate multiple values.
filter_phase	string | null	None	Filter by trial phase. Options: 'EARLY_PHASE1', 'PHASE1', 'PHASE2', 'PHASE3', 'PHASE4', 'NA'. Comma-separate multiple (e.g., 'PHASE2,PHASE3').
filter_study_type	string | null	None	Filter by study type: 'INTERVENTIONAL', 'OBSERVATIONAL', 'EXPANDED_ACCESS'.
page_size	integer	None	Number of results per page (default 10, max 1000).
next_page_token	string | null	None	Token for retrieving the next page of results (obtained from previous response).
query	string | null	None	General keyword search across all fields. Alias for query_term. E.g., "FLT3 AML", "KRAS NSCLC".
condition	string | null	None	Disease or condition to search for. Alias for query_cond. E.g., "acute myeloid leukemia". Multi-word values are matched as an exact phrase (e.g. 'kidney transplantation' is sent as "kidney transplantation"), which is narrower than a loose keyword match. For looser matching, wrap the value in parentheses (e.g. '(kidney transplantation)') to send it verbatim. Values already containing Essie syntax (AND/OR/NOT/AREA[]/quotes) are passed through unchanged. See executed_query in the response for what actually ran.
status	string | null	None	Recruitment status filter. Alias for filter_status. E.g., "RECRUITING", "COMPLETED".
max_results	integer | null	None	Maximum number of results to return. Alias for page_size.
intervention	string | null	None	Intervention or drug to search for. Alias for query_intr. E.g., "pembrolizumab", "metformin". Matched against each study's registered intervention names and synonyms (InterventionName/InterventionOtherName), not free text in study summaries, and multi-word values are matched as an EXACT PHRASE -- so 'donor derived cell therapy' matches 0 studies even though 251 trials match the same words loosely. For looser matching, wrap the value in parentheses (e.g. '(donor derived cell therapy)') to send it verbatim, or use ClinicalTrials_search_studies with query_term for free text across all study fields. Prefer the generic/INN name over a brand name. See executed_query and relaxed_match_check in the response.
sponsor	string | null	None	Lead sponsor or collaborator name to search for. E.g., "National Cancer Institute". Matched against each study's registered lead-sponsor name (AREA[LeadSponsorName]) only, and multi-word values are matched as an EXACT PHRASE. For looser matching, wrap the value in parentheses (e.g. '(National Cancer Institute)') to send it verbatim. See executed_query and relaxed_match_check in the response.
verbose	bool	True	运行提示开关。

检索 Crossref 学术文献与 DOI 元数据

dnakit.databases.crossref_search_works(*, query, limit=None, offset=None, filter=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
query	string	必填	Search query for Crossref works. Use keywords separated by spaces to refine your search across titles, abstracts, authors, and other bibliographic fields.
limit	integer	None	Number of articles to return in this page. This sets the maximum number of articles retrieved from Crossref. Max 100 per request. It does not change 'total_count', which always reports how many works match the query.
offset	integer	None	Skip this many matching works before returning results. Use with 'limit' to page through a result set larger than 100; the response 'truncation_note' gives the offset for the next page.
filter	string	None	Optional filter string for Crossref API. Examples: 'type:journal-article' (only journal articles), 'from-pub-date:2020-01-01' (published after date), 'has-abstract:true' (only with abstracts). Multiple filters comma-separated.
verbose	bool	True	运行提示开关。

检索 Europe PMC 文献与摘要

dnakit.databases.europe_pmc_search_articles(*, query, limit=None, require_has_ft=None, fulltext_terms=None, enrich_missing_abstract=None, extract_terms_from_fulltext=None, page_size=None, pageSize=None, source=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
query	string	必填	Search query for Europe PMC. Supports Lucene-like fielded syntax (e.g., BODY:"term", INTRO:"term", METHODS:"term").
limit	integer	None	Number of articles to return. This sets the maximum number of articles retrieved from Europe PMC; anything beyond it is silently left behind, so compare it against metadata.total_results in the response (Europe PMC's hitCount) before treating the returned set as complete. Europe PMC accepts up to 1000 per request.
require_has_ft	boolean	None	If true, appends HAS_FT:Y to the query to restrict results to records where Europe PMC has full text indexed. This is NOT the same as having a link to free full text elsewhere (e.g., PMC) and is often the key reason body-only keywords cannot be found via Europe PMC search.
fulltext_terms	array	None	Optional list of terms that must occur in the indexed full text (Europe PMC BODY field). When provided, the tool adds a BODY:"..." clause and automatically enables require_has_ft. {"items": {"type": "string"}}
enrich_missing_abstract	boolean	None	If true, best-effort fills missing abstracts by fetching Europe PMC fullTextXML (bounded to a few results).
extract_terms_from_fulltext	array	None	Optional list of terms to extract from full text (open access only). When provided, automatically fetches fullTextXML for open-access articles and returns snippets around these terms. Terms are processed in batches of 5 internally, so any number of terms is accepted. Bounded to first 3 OA articles to avoid latency. Returns snippets in a 'fulltext_snippets' field for each article. {"items": {"type": "string"}}
page_size	integer	None	Alias for limit — number of results to return
pageSize	integer	None	DNAKit optional alias for limit
source	string	None	数据来源筛选值；按当前接口允许的来源填写。
verbose	bool	True	运行提示开关。

按 DOI 查询 medRxiv 预印本元数据

dnakit.databases.med_rxiv_get_preprint(*, doi=None, server=None, query=None, limit=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
doi	string	None	medRxiv DOI. Can be full DOI (e.g., '10.1101/2021.04.29.21256344') or just the numeric part after '10.1101/' (e.g., '2021.04.29.21256344'). Find DOIs using EuropePMC_search_articles, web_search, or from paper citations.
server	string	None	Server to query - always 'medrxiv' for this tool.
query	string	None	检索词或查询表达式；语法由当前数据源定义。
limit	integer	None	当前请求的返回数量上限；不是全库命中总数。
verbose	bool	True	运行提示开关。

按 OpenAlex 编号查询作者详情

dnakit.databases.openalex_get_author(*, author_id, mailto=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
author_id	string	必填	OpenAlex Author ID (A...) or full OpenAlex URL.
mailto	string	None	Optional contact email for OpenAlex polite pool.
verbose	bool	True	运行提示开关。

使用 OpenAlex 检索学术文献

dnakit.databases.openalex_literature_search(*, search_keywords=None, query=None, max_results=None, limit=None, num_results=None, year_from=None, start_year=None, year_to=None, end_year=None, open_access=None, require_has_fulltext=None, fulltext_terms=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
search_keywords	string	None	Keywords to search for in paper titles/abstracts/etc. For full-text-index-only matching, also use require_has_fulltext/fulltext_terms.
query	string	None	Alias for search_keywords (recommended when you standardize on query across multiple paper-search tools).
max_results	integer	None	Maximum number of papers to retrieve (default: 10, maximum: 200).
limit	integer	None	Alias for max_results (OpenAlex max 200).
num_results	integer	None	Alias for max_results (OpenAlex max 200).
year_from	integer	None	Start year for publication date filter (e.g., 2020). Optional parameter to limit search to papers published from this year onwards.
start_year	integer	None	Alias for year_from.
year_to	integer	None	End year for publication date filter (e.g., 2023). Optional parameter to limit search to papers published up to this year.
end_year	integer	None	Alias for year_to.
open_access	boolean	None	Filter for open access papers only. Set to true for open access papers, false for non-open access, or omit for all papers.
require_has_fulltext	boolean	None	If true, filters to works where OpenAlex indicates a full-text index is available (has_fulltext:true).
fulltext_terms	array	None	Optional list of terms that must occur in OpenAlex full-text index (adds fulltext.search:<term> filters and implicitly enables require_has_fulltext). {"items": {"type": "string"}}
verbose	bool	True	运行提示开关。

以下参数组至少提供一组：search_keywords; query。

按筛选条件检索 OpenAlex 论文记录

dnakit.databases.openalex_search_works(*, search=None, query=None, filter=None, require_has_fulltext=None, fulltext_terms=None, per_page=None, limit=None, page=None, sort=None, mailto=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
search	string	None	Search query for works. Use filter + fulltext_terms/require_has_fulltext when you need full-text-index-only matching.
query	string	None	Alias for search (recommended when you standardize on query across multiple paper-search tools).
filter	string	None	OpenAlex filter string (comma-separated). Example: "from_publication_date:2020-01-01,is_oa:true".
require_has_fulltext	boolean	None	If true, appends OpenAlex filter has_fulltext:true (keeps only works with full-text index available).
fulltext_terms	array	None	Optional list of terms to match in OpenAlex full-text index. Adds one or more fulltext.search:<term> filters and implicitly enables require_has_fulltext. {"items": {"type": "string"}}
per_page	integer	None	Results per page (OpenAlex max 200).
limit	integer	None	Alias for per_page (OpenAlex max 200).
page	integer	None	Page number (1-indexed).
sort	string	None	Sort order string, e.g. "cited_by_count:desc".
mailto	string	None	Optional contact email for OpenAlex polite pool. If omitted, ToolUniverse uses a default.
verbose	bool	True	运行提示开关。

以下参数组至少提供一组：search; query。

按 PMID 获取 PubMed 文献完整元数据

dnakit.databases.pub_med_get_article(*, pmid, verbose=True)

参数	类型	必填 / 默认值	说明与约束
pmid	string | integer	必填	PubMed ID (PMID) as string or integer (e.g., '12345678' or 12345678, '33250470' or 33250470). Find PMIDs using PubMed_search_articles. Can also accept comma-separated PMIDs as string for multiple articles.
verbose	bool	True	运行提示开关。

查询引用指定 PubMed 文献的论文

dnakit.databases.pub_med_get_cited_by(*, pmid, limit=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
pmid	string	必填	PubMed ID (PMID) for which to find citing articles (e.g., '12345678'). Find PMIDs using PubMed_search_articles.
limit	integer	None	Maximum number of citing articles to return (default: 20, max: 100).
verbose	bool	True	运行提示开关。

查找与指定 PubMed 文献相似的论文

dnakit.databases.pub_med_get_related(*, pmid, limit=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
pmid	string	必填	PubMed ID (PMID) for which to find related articles (e.g., '20210808', '19879512'). Find PMIDs using PubMed_search_articles.
limit	integer	None	Maximum number of related articles to return (default: 20, max: 100).
verbose	bool	True	运行提示开关。

按检索式查询 PubMed 生物医学文献

dnakit.databases.pub_med_search_articles(*, query, limit=None, mindate=None, maxdate=None, datetype=None, include_abstract=None, sort=None, max_results=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
query	string	必填	Search query for PubMed articles. Use keywords, author names, journal names, or MeSH terms. Examples: 'cancer immunotherapy', 'Smith J[Author]', 'Nature[Journal]', 'diabetes[MeSH]'. Use AND/OR/NOT for complex queries. Note: PubMed ANDs every space-separated keyword by default, so a long, highly specific free-text query (5+ concepts) often returns 0 results even when relevant literature exists. Prefer 2-4 key concepts, add explicit OR for synonyms, or try EuropePMC_search_articles for broader relevance-ranked full-text search if this returns 0 results.
limit	integer	None	Number of articles to return. This sets the maximum number of articles retrieved from PubMed (max: 200).
mindate	string	None	Minimum date for results (format: YYYY/MM/DD or YYYY). Requires datetype to be set.
maxdate	string	None	Maximum date for results (format: YYYY/MM/DD or YYYY). Requires datetype to be set.
datetype	string	None	Type of date to filter on: 'pdat' (publication date), 'edat' (Entrez date), or 'mdat' (modification date). Required when using mindate/maxdate.
include_abstract	boolean	None	If true, best-effort fetches abstracts via efetch (adds abstract/abstract_source fields).
sort	string	None	Sort order for results. Valid values: 'pub_date' (newest first), 'Author' (alphabetical by first author), 'JournalName' (alphabetical by journal), 'relevance' (default PubMed ranking).
max_results	integer	None	Alias for limit — maximum number of results to return
verbose	bool	True	运行提示开关。

按疾病、药物或关键词查找临床试验

dnakit.databases.search_clinical_trials(*, condition=None, intervention=None, query_term=None, pageSize=None, pageToken=None, overall_status=None, max_results=None, limit=None, status=None, keyword=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
condition	string	None	Query for condition or disease using Essie expression syntax (e.g., 'lung cancer', '(head OR neck) AND pain AND NOT "back pain"'). Sent to ClinicalTrials.gov verbatim (this tool does not rewrite condition values); see executed_query in the response for what actually ran.
intervention	string	None	Query for intervention/treatment using Essie expression syntax (e.g., 'chemotherapy', 'immunotherapy', 'olaparib', 'combination therapy'). Matched against each study's registered intervention names and synonyms (InterventionName/InterventionOtherName), not free text in study summaries, so prefer the generic/INN name: a brand name that no registrant recorded in either field can return zero results. Values that already contain Essie syntax (AREA[...]/AND/OR/NOT/quotes) are passed through unchanged. Matched against each study's registered intervention names and synonyms (InterventionName/InterventionOtherName), not free text in study summaries, and multi-word values are matched as an EXACT PHRASE -- so 'donor derived cell therapy' matches 0 studies even though 251 trials match the same words loosely. For looser matching, wrap the value in parentheses (e.g. '(donor derived cell therapy)') to send it verbatim, or use ClinicalTrials_search_studies with query_term for free text across all study fields. Prefer the generic/INN name over a brand name. See executed_query and relaxed_match_check in the response.
query_term	string	None	Query for 'other terms' with Essie expression syntax (e.g., 'combination', 'AREA[LastUpdatePostDate]RANGE[2023-01-15,MAX]', 'Phase II'). Can be used to search for all other protocol fields, including but not limited to title, outcome measures, status, phase, location, etc.
pageSize	integer	None	Maximum number of studies to return per page (default 10, max 1000).
pageToken	string	None	Token to retrieve the next page of results, obtained from the 'nextPageToken' field of the previous response. Do not specify it for first page. When you make an initial request to the API which supports pagination, the response will include a nextPageToken. This token can then be used as a parameter in the subsequent API request to retrieve the next set of data.
overall_status	array	None	Filter by overall study status (e.g., ['RECRUITING'], ['COMPLETED'], ['RECRUITING', 'NOT_YET_RECRUITING']). Valid values: RECRUITING, NOT_YET_RECRUITING, ACTIVE_NOT_RECRUITING, COMPLETED, ENROLLING_BY_INVITATION, SUSPENDED, TERMINATED, WITHDRAWN. {"items": {"type": "string"}}
max_results	integer	None	Maximum number of studies to return (alias for pageSize, default 10, max 1000).
limit	integer	None	Alias for max_results: maximum number of studies to return (default 10, max 1000).
status	object | None	None	Alias for overall_status. Filter by trial status, e.g. "RECRUITING", "COMPLETED". {"oneOf": [{"type": "string"}, {"type": "array", "items": {"type": "string"}}]}
keyword	string	None	Alias for query_term. Free-text keyword search across all trial fields (e.g., drug name, condition, investigator).
verbose	bool	True	运行提示开关。

检索论文、摘要与简短总结

dnakit.databases.semantic_scholar_search_papers(*, query, limit=None, year=None, sort=None, include_abstract=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
query	string	必填	Search query for Semantic Scholar. Use keywords separated by spaces to refine the search.
limit	integer	None	Maximum number of papers to return from Semantic Scholar.
year	string	None	Filter results by publication year. Use a single year (e.g., '2024') or a range (e.g., '2020-2024').
sort	string	None	Sort results. Options: 'citationCount:desc', 'citationCount:asc', 'publicationDate:desc', 'publicationDate:asc'.
include_abstract	boolean	None	If true, best-effort fetches missing abstracts via the paper detail endpoint (only when abstract is missing in search results).
verbose	bool	True	运行提示开关。

通过搜索引擎检索网页

dnakit.databases.web_search(*, query, max_results=None, search_type=None, backend=None, region=None, safesearch=None, verbose=True)

参数	类型	必填 / 默认值	说明与约束
query	string	必填	Search query string
max_results	integer	None	Maximum number of results to return
search_type	string	None	Type of search to perform
backend	string	None	Search engine backend to use. Selecting 'parallel' sends the query to https://search.parallel.ai/mcp; that backend does not apply ToolUniverse region or safesearch controls. Selecting 'serpbase' sends the query to https://api.serpbase.dev (requires SERPBASE_API_KEY) for real Google SERP results; that backend applies region but not safesearch.
region	string	None	Search region/locale for DDGS backends, and country/language for the SerpBase backend (split on the hyphen into gl/hl). Parallel does not support this control; only the default value is accepted with backend='parallel'.
safesearch	string	None	Safe search level for DDGS backends. Parallel and SerpBase do not support this control; only the default value is accepted with those backends.
verbose	bool	True	运行提示开关。