0.4 起旧名不再导出。查询替换名：
library(bulkMAE)
bulkmae_rename("run_gsva")
#> [1] "score_gsva()"
bulkmae_rename("run_go_ora")
#> [1] "enrich_go(method = \"ora\")"完整对照表见随包指南
system.file("guides", "naming-migration-zh.md", package = "bulkMAE")。
0.4.0 将公开 API 统一为 <family>_<method>()
或
<family>_<operation>()。这是有意的破坏性重命名：旧函数不再导出，也不保留
deprecated aliases，避免自动补全继续出现
run_*、infer_*、prepare_*
和新函数并存的情况。
命名边界
-
mae_*：MAE/SE 构建、校验和取值。 -
qc_*、filter_*、reduce_*：质控、过滤和样本降维。 -
normalize_*、transform_*、adjust_*：严格区分归一化、尺度变换和批次/ 隐变量校正。 -
de_*、dtu_*：表达模型/差异分析与转录本使用。 -
enrich_*、score_*、activity_*：集合检验、样本级评分和调控活性。 -
cluster_*、coexpr_*、network_*：分型、共表达和一般网络。 -
deconv_*、surv_*、ml_*、meta_*、drug_*：组织反卷积与临床转化。
本版不额外增加只有 ... 的总调度器 de() 或
enrich()。不同后端对
design、formula、contrast、输入尺度和返回对象的要求差异很大；保留
可检查签名的薄函数更符合无状态、不过度设计的边界。deconv()
是例外，因为 它直接对应 immunedeconv 已有的统一方法契约。
旧名到新名
| 区域 | 0.3 及更早 | 0.4.0 |
|---|---|---|
| MAE | validate_mae() |
mae_validate() |
| MAE | experiment_names() |
mae_experiments() |
| MAE | assay_names() |
mae_assays() |
| MAE | pull_experiment() |
mae_pull_experiment() |
| MAE | pull_assay() |
mae_pull_assay() |
| MAE | sample_data() |
mae_samples() |
| MAE | make_experiment() |
mae_create_experiment() |
| MAE | make_mae() |
mae_create() |
| Annotation | strip_ensembl_version() |
annotate_ensembl() |
| Annotation | map_feature_ids() |
annotate_ids() |
| Annotation | query_biomart() |
annotate_biomart() |
| QC | filter_low_expression() |
filter_expr() |
| QC | library_metrics() |
qc_library() |
| QC | sample_correlations() |
qc_correlation() |
| Reduction | run_pca() |
reduce_pca() |
| Reduction | run_mds() |
reduce_mds() |
| Reduction | run_umap() |
reduce_umap() |
| Reduction | run_tsne() |
reduce_tsne() |
| QC | flag_pca_outliers() |
qc_outliers() |
| Normalize | estimate_size_factors() |
normalize_deseq() |
| Adjust | remove_batch_effect() |
adjust_batch() |
| Adjust | adjust_combat_seq() |
adjust_combatseq() |
| Adjust | estimate_sva() |
adjust_sva() |
| Adjust | run_ruvg() |
adjust_ruv() |
| Model | partition_variance() |
de_variance() |
| Differential | run_deseq2() |
de_deseq2() |
| Differential | deseq2_results() |
de_deseq2_results() |
| Differential | run_edger() |
de_edger() |
| Differential | run_limma_voom() |
de_voom() |
| Differential | run_limma() |
de_limma() |
| DTU | run_diff_splice() |
dtu_diffsplice() |
| Differential | run_masigpro() |
de_masigpro() |
| Differential | run_dream() |
de_dream() |
| Enrichment | run_ora() |
enrich_ora() |
| Enrichment | run_go_ora() |
enrich_go(method = "ora") |
| Enrichment | run_goseq() |
enrich_goseq() |
| Enrichment | run_kegg_ora() |
enrich_kegg(method = "ora") |
| Enrichment | run_reactome_ora() |
enrich_reactome(method = "ora") |
| Enrichment | run_fgsea() |
enrich_fgsea() |
| Enrichment | run_gene_set_gsea() |
enrich_gsea() |
| Enrichment | run_kegg_gsea() |
enrich_kegg(method = "gsea") |
| Enrichment | run_reactome_gsea() |
enrich_reactome(method = "gsea") |
| Enrichment | run_camera() |
enrich_camera() |
| Enrichment | run_fry() |
enrich_fry() |
| Enrichment | run_mroast() |
enrich_roast() |
| Score | run_gsva() |
score_gsva() |
| Score | run_ssgsea() |
score_ssgsea() |
| Score | run_singscore() |
score_singscore() |
| Activity | decouple_methods() |
activity_methods() |
| Activity | get_decouple_resource() |
activity_resource() |
| Activity | run_decouple() |
activity_decouple() |
| Activity | infer_progeny() |
activity_progeny() |
| Activity | infer_tf_activity() |
activity_tf() |
| Cluster | run_consensus_clustering() |
cluster_consensus() |
| Cluster | run_consensus_icl() |
cluster_consensus_diagnostics() |
| Cluster | consensus_classes() |
cluster_consensus_classes() |
| Co-expression | run_differential_coexpression() |
coexpr_differential() |
| Cluster | run_nmf() |
cluster_nmf() |
| Co-expression | run_wgcna() |
coexpr_wgcna() |
| Co-expression | run_module_preservation() |
coexpr_preservation() |
| Network | run_genie3() |
network_genie3() |
| Network | genie3_links() |
network_genie3_links() |
| Network | run_string_network() |
network_string() |
| Deconvolution | run_immunedeconv() |
deconv() |
| Deconvolution | run_music() |
deconv_music() |
| Deconvolution | prepare_cibersortx_input() |
deconv_cibersortx_input() |
| Deconvolution | run_bayesprism() |
deconv_bayesprism() |
| Survival | run_kaplan_meier() |
surv_km() |
| Survival | run_time_roc() |
surv_roc() |
| Survival | run_cox() |
surv_cox() |
| Survival | run_penalized_cox() |
surv_penalized() |
| Machine learning | run_classifier_glmnet() |
ml_glmnet() |
| Meta-analysis | run_meta_analysis() |
meta_effect() |
| Drug | prepare_lincs_query() |
drug_query() |
| Drug | run_lincs_search() |
drug_lincs() |
保持不变的名称
import_tximport()、normalize_tmm()、transform_voom()、
transform_vst()、transform_rlog()、adjust_combat()
和 score_signature() 已符合新规则，不需重命名。
数据库富集的合并
GO、KEGG 和 Reactome 不再为 ORA/GSEA 建立组合式函数名。使用：
enrich_go(genes = selected, method = "ora", org_db = org.Hs.eg.db)
enrich_go(ranks = statistic, method = "gsea", org_db = org.Hs.eg.db)
enrich_kegg(genes = selected, method = "ora")
enrich_kegg(ranks = statistic, method = "gsea")
enrich_reactome(genes = selected, method = "ora")
enrich_reactome(ranks = statistic, method = "gsea")这样将“统计方法”保留为
method，将“知识库”保留在函数家族名中， 避免
ora_* 和 gsea_* 组合随后端增长。