Skip to content

MIxS schema diff: v5 → v6.0.0

v5 (600 terms, from mixs-legacy mixs_v5.xlsx) → v6.0.0 (804 terms, from GenomicsStandardsConsortium/mixs@mixs6.0.0). Headline counts: 212 added, 20 renamed, 4 removed, 3 deleted, 1 rename candidate, 65 definitions changed, 573 shared.

The 204-term net growth is real expansion, not churn: almost all of it is new measurement packages (Food, host-of-host symbiosis, animal husbandry, spike-in controls, fermentation). The removals are small and mostly structural.

Added (212), by theme

The additions cluster into a handful of new or heavily expanded domains rather than scattered one-off fields.

  • Food package (largest single driver). 31 slots beginning food_ (packaging, processing, distribution, provenance, traceability: e.g. food_pack_integrity, food_dis_point, food_trace_list, food_treat_proc), plus the food-adjacent fields cons_food_stor_dur, cons_food_stor_temp, cons_purch_date, cons_qty_purchased, dietary_claim_use, intended_consumer, spec_intended_cons, prod_label_claims, HACCP_term, IFSAC_category, facility_type, hygienic_area, env_monitoring_zone, lot_number, ster_meth_samp_room, and the enums Food_Product_type / Food_source. This is an entire new package arriving at once.
  • Host-of-host / symbiosis (new domain). 14 host_of_host_* slots (taxid, name, disease, genotype, phenotype, gravidity, total mass, environment, etc.) plus host_symbiont, host_specificity, host_dependence, sym_life_cycle_type, symbiont_host_role, type_of_symbiosis, mode_transmission, route_transmission, association_duration. MIxS gained the ability to describe a host that itself lives on/in another host.
  • Animal husbandry / farm. 12 animal_* slots (antimicrobial use animal_am*, animal_housing, animal_diet, animal_body_cond, ...), the farm_equip* cluster (4), farm_water_source, anim_water_method, fertilizer_admin, fertilizer_date, crop_yield, pres_animal_insect.
  • Spike-in controls. 8 spikein_* slots (spikein_AMR, spikein_antibiotic, spikein_org, spikein_serovar, spikein_strain, ...) plus neg_cont_type, pos_cont_type.
  • Fermentation. 9 ferm_* slots (ferm_pH, ferm_temp, ferm_time, ferm_vessel, ferm_headspace_oxy, ...).
  • Microbial culture / enrichment. microb_start* (5), microb_cult_med, micro_biomass_meth, cult_isol_date, cult_result, cult_result_org, cult_target, growth_medium, enrichment_protocol, bacterial_density, serovar_or_serotype.
  • Sample handling, storage, transport. samp_stor_* (5: device/dur/loc/media/temp) + samp_store_sol, samp_transport_* (3), samp_rep_biol/samp_rep_tech, samp_pooling, samp_purpose, samp_source_mat_cat, samp_taxon_id, samp_surf_moisture, samp_loc_condition, area_samp_size, num_samp_collect.
  • Soil (expanded). 10 soil_* slots (soil_pH, soil_horizon, soil_porosity, soil_texture_class, soil_texture_meth, ...), plus part_plant_animal, plant_part_maturity, plant_reprod_crop, plant_water_method, photosynt_activ(+_meth), prev_land_use_meth, non_min_nutr_regm.
  • Sequencing / library prep. library_prep_kit, nucl_acid_ext_kit, sequencing_kit, sequencing_location.
  • Study design. study_design, study_inc_dur, study_inc_temp, study_timecourse, study_tmnt, timepoint, biocide_used.
  • Air / weather / location context. air_PM_concen, air_flow_impede, adjacent_environment, coll_site_geo_feat, extr_weather_event, date_extr_weath, season_humidity, rel_location, water_source_adjac, water_source_shared, water_pH, water_frequency.
  • Nutrient chemistry. tot_car, tot_phos, tot_n_meth.
  • LinkML infrastructure (not measurement slots). Several added names are the schema's own scaffolding, not sample metadata: core field, environment field, investigation field, sequencing field, nucleic acid sequence source field, mixs extension field, associated resource, has numeric value, has raw value, has unit, repository_name, associated resource. These reflect the v6 move to a normalized LinkML model with mixin "field" classes and a value/unit/raw-value pattern. Count them separately from the ~200 real new slots.

Removed (4) and deleted (3)

The diff distinguishes two removal categories:

  • removed (4): extreme_salinity, nose_mouth_teeth_throat_disord, pres_animal, resp_part_matter.
  • deleted (3): env_package, investigation_type, submitted_to_insdc. These three are administrative/structural fields, dropped as part of the v6 model reorganization rather than retired measurements.

Two of the four removed slots look like replacements rather than losses: pres_animal is the rename candidate below (pres_animal_insect was added), and resp_part_matter (respirable particulate matter) is plausibly superseded by the new air-particulate fields air_PM_concen / air particulate matter concentration. nose_mouth_teeth_throat_disord overlaps the surviving nose_throat_disord. None of these are stated as renames, so I have not treated them as such beyond flagging.

Renamed (20), old → new

Three consistent conventions explain nearly all 20 renames:

  1. Abbreviation truncation (the dominant pattern): treatment→treat, system→sys, content→cont, production→prod, collection→collect, element→elem, damage→dam, device→dev, shading→shad, pressure→press, organism→org.
  2. chem_treatment_method→chem_treat_method, heat_system_deliv_meth→heat_sys_deliv_meth, tot_nitro_content_meth→tot_nitro_cont_meth, water_content_soil_meth→water_cont_soil_meth, water_production_rate→water_prod_rate, samp_collection_point→samp_collect_point, room_architec_element→room_architec_elem, room_moist_damage_hist→room_moist_dam_hist, shading_device_water_mold→shad_dev_water_mold, tvdss_of_hcr_pressure→tvdss_of_hcr_press, organism_count_qpcr_info→org_count_qpcr_info.
  3. vOTU → OTU terminology change (drop the leading v): votu_class_appr→otu_class_appr, votu_db→otu_db, votu_seq_comp_appr→otu_seq_comp_appr. (The matching definitions also change vOTU→OTU wording, see below.)
  4. Leading x_ added to digit-initial names (LinkML identifiers cannot start with a digit): 16s_recover→x_16s_recover, 16s_recover_software→x_16s_recover_software.

The remaining renames drop a semantic prefix: host_blood_press_diast→blood_press_diast, host_blood_press_syst→blood_press_syst, ihmc_ethnicity→ethnicity, and health_disease_stat→host_disease_stat (a health→host reframing that is also a definition change; see Notes).

Possible missed renames

  • pres_animal → pres_animal_insect (1 candidate). pres_animal appears in the removed list and pres_animal_insect in added. This is almost certainly a real rename that a maintainer should confirm and promote into the tool's rename map, so it stops being reported as a removal + an unrelated addition.

Definition changes (65)

Two cross-cutting drivers account for most of the substantive edits; a third large group is purely cosmetic.

Driver 1: adding controlled-vocabulary / ontology references. The clearest single theme. Roughly ten disease-history fields all gained a Human Disease Ontology (DO) reference they lacked in v5: blood_blood_disord, dermatology_disord, gastrointest_disord, gynecologic_disord, kidney_disord, liver_disord, nose_throat_disord, pulmonary_disord, urogenit_disord, urogenit_tract_disor. Beyond disease fields, many slots shifted from a prose definition to pointing at an OBO term list: samp_collect_device (ENVO + GENEPIO), samp_type (GENEPIO), soil_type (ENVO), season (NCIT), seq_meth (OBI DNA-sequencer list), and the env_* triad env_broad_scale / env_local_scale / env_medium (EnvO usage-wiki links). v6 is systematically wiring free-text fields to ontologies.

Driver 2: concept → tool/method, and genuine redefinitions. A cluster of fields moved from describing a concept to naming the tool or measurement: - chimera_check: v5 defined what a chimera is; v6 makes it the tool used for chimera checking. - seq_meth: "sequencing method" → "sequencing machine" (OBI list). - bin_software: now asks for a product ID.

Genuine meaning changes that could affect existing data (worth individual attention): - host_spec_range: v5 = "NCBI taxid of the specific host"; v6 = "the range and diversity of host species an organism can infect." The field's meaning was effectively inverted. - health_disease_stat→host_disease_stat: v5 = "health or disease status"; v6 = "list of diseases diagnosed." Rename + redefinition together. - propagation: expanded and re-scoped (phage/virus/plasmid/eukaryote reproduction types). - specific_host: narrowed to "host's taxonomic name and/or NCBI taxonomy ID." - Generalizations: compl_appr (SAG/MAG → any genomic assembly), salinity (water → liquid or solid), freq_clean (building/week → any sample location, any cadence), samp_size / samp_vol_we_dna_ext (units clarified), depth (v5 punted to package docs; v6 gives a real definition), host_sex ("physical sex" → "gender or physical sex"). - url: v5 description was empty; v6 supplies one (previously-undocumented field now documented). - votu_class_appr / votu_db / votu_seq_comp_appr: definitions change vOTU→OTU to match the renames above.

Cosmetic changes (grouped). Around 20 of the 65 entries are text-only edits with no change of meaning: - Mojibake / encoding cleanup (v5 had ’,  , ° artifacts from the xlsx export): alt, pour_point, viscosity, and others where the only diff is a curly apostrophe or non-breaking space becoming clean text. - Double-space collapsed to single space: chem_treatment, fungicide_regm, hcr, iw_bt_date_well, pcr_cond, samp_collection_point, vir_ident_software, and more. - Trailing period added: biol_stat, elev, host_family_relationship, host_infra_specific_name, host_substrate, ref_biomaterial, temp, texture_meth, tvdss_of_hcr_temp, tvdss_of_hcr_pressure.

These are reported as three grouped mass-edits, not ~20 separate changes.

Notes (judgment calls and data-quality flags)

  • Near-duplicate new slot names. The added list contains what look like three spellings of one concept: samp_collec_method, samp_collect_method, and sample_collec_method; plus two "sample name" spellings: samp_name and sample_name; and samp_collec_device alongside them. These are probably the same measurement named inconsistently across packages. A maintainer should confirm whether they are intentional distinct fields or accidental variants.
  • Copy-paste bug in gastrointest_disord. Its v6 definition contains a stray "History of blood disorders; can include multiple disorders." pasted in from blood_blood_disord. Introduced in v6, worth a fix upstream.
  • health_disease_stat / host_disease_stat double-accounting. health_disease_stat→host_disease_stat appears as a rename, and both names also appear under definition_changed. Reading it as a single rename-plus-redefinition (not two independent slot edits) avoids overcounting.
  • geo_loc_name dropped the GAZ ontology version pin "(v 1.512)"; minor content deletion, not cosmetic.
  • No cardinality, range, or pattern data is present in this diff shape, so those sections are omitted. If those changed between v5 and v6, they are not captured here.