microbiomedata / mixs-6-2-release-candidate Goto Github PK

mine and aggregate other issues here

Include ambiguity codes?

^[ACGTRKSYMWBHDVN]+$

Constrain the prefix with a pattern

Separation of concerns

cleanup for all strings, whether they are global or part of a slot_usage should follow the same pattern

Could the user possibly think "I should enter a code for that thing!"

If so, move the Expected value annotation into another slot property, like description.

see also

• #36

generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'api' is a required property
generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'basin' is a required property
generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'coll_site_geo_feat' is a required property
generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'collection_date' is a required property
generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'env_broad_scale' is a required property
generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'env_local_scale' is a required property
generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'env_medium' is a required property
generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'iwf' is a required property
generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'microbial_biomass_meth' is a required property
generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'occup_density_samp' is a required property
generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0]: 'water_cut' is a required property

• generated/GSC_MIxS_6_usage_populated_no_blank_cols.tsv
• which is generated by generate_and_populate_template in project.Makefile
• that doesn't retain extra schemasheet header rows, for annotation inner_keys, internal_separators (for lists)

• infer enumerations as long as string_serialization starts with [, contains |s and doesn't include , : or (
• make sure examples from slots using the enum match a PV form the enum
• #55
• #56

I don't think I ever completed this. Will check if I have the branch locally.

For all individual checklist and environmental package classes, or just combos?

also body mass index BMI

https://github.com/only1chunts/mixs-cih-fork/raw/main/mixs/excel/mixs_v6.xlsx

linklml-validate will report all usages of undefined slots, instead of failing with a stack trace on the first one

GSC/MIxS TWG people have is said that any slot from any checklist or environmental package can be combined together in any submission, and even that submissions with non-MIxS slots should be accepted

The problem with this is that submitter may think of their own term names/column heading for a variable, not realizing that MIxS already has a term with the same semantics. That reduces the FAIRness of the data.

@only1chunts suggested rating submissions with 0 to 5 stars

I will experiment some more with extra slots in closed and open schemas

see also

• #117

I think they mean |-separated strings, which require more sophisticated regular expressions

• MIXS:0000018: depth & soil_depth
• MIXS:0000091: associated_resource & url
• MIXS:0000110: samp_stor_temp & samp_store_temp
• MIXS:0000116: samp_stor_dur & samp_store_dur
• MIXS:0000335: soil_text_measure & texture
• MIXS:0000336: soil_texture_meth & texture_meth
• MIXS:0000338: tot_n_meth & tot_nitro_cont_meth
• MIXS:0000525: tot_car & tot_carb
• MIXS:0000689: tot_phos & tot_phosphate
• MIXS:0000755: samp_stor_loc & samp_store_loc
• MIXS:0001107: samp_name & sample_name
• MIXS:0001139: Food_source & food_sourc
• MIXS:0001184: Food_Product_type & food_product_type
• MIXS:0001225: samp_collect_method & sample_collec_method

see also

• #13

• checklists should be mixins
  • try that with environmental package classes too?
• addition of combinations for all checklists X all environmental packages will make the schema painfully large

see also

• #8

generated/mixs_v6.xlsx.examples.yaml::$.exhaustive_test_set[0].time_last_toothbrush: 'P2H45M' does not match '^P(?!$)(\d+Y)?(\d+M)?(\d+W)?(\d+D)?(T(?=\d+[HMS])(\d+H)?(\d+M)?(\d+S)?)?$'

• automate
• rules for when to trust the example vs when to trust the Value syntax, Expected value, etc?
• some provided examples might actually contain multiple examples, potentially split with multiple different delimiters

check these repos:

• https://github.com/GenomicsStandardsConsortium/mixs
  • https://github.com/GenomicsStandardsConsortium/mixs/blob/issue-511-tested-schemasheets/schemasheets/tsv_in/MIxS_6_term_updates_classdefs.tsv
• https://github.com/microbiomedata/mixs_fresh_start
• https://github.com/turbomam/mixs-subset-examples-first

See https://github.com/turbomam/mixs-envo-struct-knowl-extraction/blob/main/data/string_ser_exp_val_to_range_pattern.tsv

Note: will be updating file name to remove whitespace

• would only support a single unit
• might be hard to work with in schemasheets

from https://github.com/GenomicsStandardsConsortium/mixs/wiki/5.-MIxS-checklists

• not applicable (-): descriptor is not applicable for a given checklist type

  • implemented in LinkML by deleting an association between the named term on a row in the XLSX sheet and the class implied

• C conditional mandatory (C): descriptor must be present for compliance with the checklist, but only when applicable to the study, i.e. if this item is not applicable for the study the metadata data will still be checklist compliant even if it is left out

  • implemented in LinkML as recommended: true

• E environment-dependent (E): descriptor must be present depending on the environment the original sample was obtained from

  • globally implemented as recommended: true
  • in the slot usage of the named term and the implied class, implemented as required: true

• M mandatory (M): descriptor must be present for compliance with the checklist_

  • implemented in LinkML as required: true

• X optional (X): descriptor may be present, not mandatory for compliance with checklist

  • implemented in LinkML as required: false and recommended: false

• #18
• #19
• #20
• #21

and might need to be split up