conseil scientifique metaseed : axe 4

Il existe dans le monde trois grandes bases de données susceptibles de nous intéresser : SRA (the NCBI's Sequence Read Archive -- NGS data), ENA (the EMBL's European Nucleotide Archive), DBA (the DDBJ's Sequence Read Archive).

Malheureusement pour nous, ces bases sont plus prévues pour le dépot de séquences que pour une recherche «intelligente« de données. De plus ces bases sont sans doute miroirs les unes des autres, mais avec leur propre système de numérotation des study, run, experiment.

Comme souvent dans une base de données, le plus simple -- pour un(e) informaticien(ne) -- est de connaitre l'identifiant de ce qu'on cherche, ce qui bien sûr est une tache difficile voire ingrate pour un utiisateur de base. Cliquer sur l'image ci-dessous pour voir comment on navigue dans SRA :

Une première tentative, initiée dans le cadre du stage de M2 de QuRP, a été de rechercher des mots-clés dans une vingtaine de publis puis d'interroger les bases de données avec ces mots-clés. Cette approche est moyennement satisfaisante dans la mesure où il y avait beaucoup de faux positifs renvoyés avec ces termes.

Nous avons donc tenté une autre approche qui consiste à trouver directement ces numéros d'accession, comme PRJNA208116, PRJEB5860 ou SRR1524567 dans des articles sélectionnés par MaBA. Voici deux des articles choisis :

Bulgarelli 2015	Structure and function of the bacterial root microbiota in...
Knief 2012	Metaproteogenomic analysis of microbial communities in the phyllosphere and...

Pour éviter de passer trop de temps à lire nous-mêmes les articles (ce qui s'automatise mal), nous avons essayé de détecter l'environnement de mots comme database, accession, number. Et effectivement, il est assez simple d'extraire des articles des informations comme

Une fois ces numéros d'accession trouvés (ou pas!), il faut savoir à quelles données ils correspondent. Heureusement, le logiciel R fournit un package nommé SRAdb dont la prise en main est assez facile... une fois les descripteurs de la base chargée, ce qui prend un certain temps car il faut rapatrier 15 Go de métadonnées. Les champs (ou colonnes) de la base de données SRA ont des noms explicites comme on peut le voir ci-dessous :

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+                                                         +
+ Liste des sra_tables (tables de données dans SRA)       +
+                                                         +
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

[1,] "col_desc"
 [2,] "experiment"
 [3,] "fastq"
 [4,] "metaInfo"
 [5,] "run"
 [6,] "sample"
 [7,] "sra"
 [8,] "sra_ft"
 [9,] "sra_ft_content"
[10,] "sra_ft_segdir"
[11,] "sra_ft_segments"
[12,] "study"
[13,] "submission"

table col_desc
==============

[,1]
[1,] "col_desc_ID"
[2,] "table_name"
[3,] "field_name"
[4,] "type"
[5,] "description"
[6,] "value_list"
[7,] "sradb_updated"

table experiment
================

[,1]
 [1,] "experiment_ID"
 [2,] "bamFile"
 [3,] "fastqFTP"
 [4,] "experiment_alias"
 [5,] "experiment_accession"
 [6,] "broker_name"
 [7,] "center_name"
 [8,] "title"
 [9,] "study_name"
[10,] "study_accession"
[11,] "design_description"
[12,] "sample_name"
[13,] "sample_accession"
[14,] "sample_member"
[15,] "library_name"
[16,] "library_strategy"
[17,] "library_source"
[18,] "library_selection"
[19,] "library_layout"
[20,] "targeted_loci"
[21,] "library_construction_protocol"
[22,] "spot_length"
[23,] "adapter_spec"
[24,] "read_spec"
[25,] "platform"
[26,] "instrument_model"
[27,] "platform_parameters"
[28,] "sequence_space"
[29,] "base_caller"
[30,] "quality_scorer"
[31,] "number_of_levels"
[32,] "multiplier"
[33,] "qtype"
[34,] "sra_link"
[35,] "experiment_url_link"
[36,] "xref_link"
[37,] "experiment_entrez_link"
[38,] "ddbj_link"
[39,] "ena_link"
[40,] "experiment_attribute"
[41,] "submission_accession"
[42,] "sradb_updated"

table fastq
===========

[,1]
[1,] "fastq_ID"
[2,] "run_accession"
[3,] "file_name"
[4,] "md5"
[5,] "bytes"
[6,] "audit_time"
[7,] "sradb_updated"

table metaInfo
==============

[,1]
[1,] "name"
[2,] "value"

table run
=========

[,1]
 [1,] "run_ID"
 [2,] "bamFile"
 [3,] "run_alias"
 [4,] "run_accession"
 [5,] "broker_name"
 [6,] "instrument_name"
 [7,] "run_date"
 [8,] "run_file"
 [9,] "run_center"
[10,] "total_data_blocks"
[11,] "experiment_accession"
[12,] "experiment_name"
[13,] "sra_link"
[14,] "run_url_link"
[15,] "xref_link"
[16,] "run_entrez_link"
[17,] "ddbj_link"
[18,] "ena_link"
[19,] "run_attribute"
[20,] "submission_accession"
[21,] "sradb_updated"

table sample
============

[,1]
 [1,] "sample_ID"
 [2,] "sample_alias"
 [3,] "sample_accession"
 [4,] "broker_name"
 [5,] "center_name"
 [6,] "taxon_id"
 [7,] "scientific_name"
 [8,] "common_name"
 [9,] "anonymized_name"
[10,] "individual_name"
[11,] "description"
[12,] "sra_link"
[13,] "sample_url_link"
[14,] "xref_link"
[15,] "sample_entrez_link"
[16,] "ddbj_link"
[17,] "ena_link"
[18,] "sample_attribute"
[19,] "submission_accession"
[20,] "sradb_updated"

table sra
=========

[,1]
 [1,] "sra_ID"
 [2,] "SRR_bamFile"
 [3,] "SRX_bamFile"
 [4,] "SRX_fastqFTP"
 [5,] "run_ID"
 [6,] "run_alias"
 [7,] "run_accession"
 [8,] "run_date"
 [9,] "updated_date"
[10,] "spots"
[11,] "bases"
[12,] "run_center"
[13,] "experiment_name"
[14,] "run_url_link"
[15,] "run_entrez_link"
[16,] "run_attribute"
[17,] "experiment_ID"
[18,] "experiment_alias"
[19,] "experiment_accession"
[20,] "experiment_title"
[21,] "study_name"
[22,] "sample_name"
[23,] "design_description"
[24,] "library_name"
[25,] "library_strategy"
[26,] "library_source"
[27,] "library_selection"
[28,] "library_layout"
[29,] "library_construction_protocol"
[30,] "adapter_spec"
[31,] "read_spec"
[32,] "platform"
[33,] "instrument_model"
[34,] "instrument_name"
[35,] "platform_parameters"
[36,] "sequence_space"
[37,] "base_caller"
[38,] "quality_scorer"
[39,] "number_of_levels"
[40,] "multiplier"
[41,] "qtype"
[42,] "experiment_url_link"
[43,] "experiment_entrez_link"
[44,] "experiment_attribute"
[45,] "sample_ID"
[46,] "sample_alias"
[47,] "sample_accession"
[48,] "taxon_id"
[49,] "common_name"
[50,] "anonymized_name"
[51,] "individual_name"
[52,] "description"
[53,] "sample_url_link"
[54,] "sample_entrez_link"
[55,] "sample_attribute"
[56,] "study_ID"
[57,] "study_alias"
[58,] "study_accession"
[59,] "study_title"
[60,] "study_type"
[61,] "study_abstract"
[62,] "center_project_name"
[63,] "study_description"
[64,] "study_url_link"
[65,] "study_entrez_link"
[66,] "study_attribute"
[67,] "related_studies"
[68,] "primary_study"
[69,] "submission_ID"
[70,] "submission_accession"
[71,] "submission_comment"
[72,] "submission_center"
[73,] "submission_lab"
[74,] "submission_date"
[75,] "sradb_updated"

table sra_ft
============

[,1]
 [1,] "SRR_bamFile"
 [2,] "SRX_bamFile"
 [3,] "SRX_fastqFTP"
 [4,] "run_ID"
 [5,] "run_alias"
 [6,] "run_accession"
 [7,] "run_date"
 [8,] "updated_date"
 [9,] "spots"
[10,] "bases"
[11,] "run_center"
[12,] "experiment_name"
[13,] "run_url_link"
[14,] "run_entrez_link"
[15,] "run_attribute"
[16,] "experiment_ID"
[17,] "experiment_alias"
[18,] "experiment_accession"
[19,] "experiment_title"
[20,] "study_name"
[21,] "sample_name"
[22,] "design_description"
[23,] "library_name"
[24,] "library_strategy"
[25,] "library_source"
[26,] "library_selection"
[27,] "library_layout"
[28,] "library_construction_protocol"
[29,] "adapter_spec"
[30,] "read_spec"
[31,] "platform"
[32,] "instrument_model"
[33,] "instrument_name"
[34,] "platform_parameters"
[35,] "sequence_space"
[36,] "base_caller"
[37,] "quality_scorer"
[38,] "number_of_levels"
[39,] "multiplier"
[40,] "qtype"
[41,] "experiment_url_link"
[42,] "experiment_entrez_link"
[43,] "experiment_attribute"
[44,] "sample_ID"
[45,] "sample_alias"
[46,] "sample_accession"
[47,] "taxon_id"
[48,] "common_name"
[49,] "anonymized_name"
[50,] "individual_name"
[51,] "description"
[52,] "sample_url_link"
[53,] "sample_entrez_link"
[54,] "sample_attribute"
[55,] "study_ID"
[56,] "study_alias"
[57,] "study_accession"
[58,] "study_title"
[59,] "study_type"
[60,] "study_abstract"
[61,] "center_project_name"
[62,] "study_description"
[63,] "study_url_link"
[64,] "study_entrez_link"
[65,] "study_attribute"
[66,] "related_studies"
[67,] "primary_study"
[68,] "submission_ID"
[69,] "submission_accession"
[70,] "submission_comment"
[71,] "submission_center"
[72,] "submission_lab"
[73,] "submission_date"
[74,] "sradb_updated"

table sra_ft_content
====================

[,1]
 [1,] "docid"
 [2,] "c0SRR_bamFile"
 [3,] "c1SRX_bamFile"
 [4,] "c2SRX_fastqFTP"
 [5,] "c3run_ID"
 [6,] "c4run_alias"
 [7,] "c5run_accession"
 [8,] "c6run_date"
 [9,] "c7updated_date"
[10,] "c8spots"
[11,] "c9bases"
[12,] "c10run_center"
[13,] "c11experiment_name"
[14,] "c12run_url_link"
[15,] "c13run_entrez_link"
[16,] "c14run_attribute"
[17,] "c15experiment_ID"
[18,] "c16experiment_alias"
[19,] "c17experiment_accession"
[20,] "c18experiment_title"
[21,] "c19study_name"
[22,] "c20sample_name"
[23,] "c21design_description"
[24,] "c22library_name"
[25,] "c23library_strategy"
[26,] "c24library_source"
[27,] "c25library_selection"
[28,] "c26library_layout"
[29,] "c27library_construction_protocol"
[30,] "c28adapter_spec"
[31,] "c29read_spec"
[32,] "c30platform"
[33,] "c31instrument_model"
[34,] "c32instrument_name"
[35,] "c33platform_parameters"
[36,] "c34sequence_space"
[37,] "c35base_caller"
[38,] "c36quality_scorer"
[39,] "c37number_of_levels"
[40,] "c38multiplier"
[41,] "c39qtype"
[42,] "c40experiment_url_link"
[43,] "c41experiment_entrez_link"
[44,] "c42experiment_attribute"
[45,] "c43sample_ID"
[46,] "c44sample_alias"
[47,] "c45sample_accession"
[48,] "c46taxon_id"
[49,] "c47common_name"
[50,] "c48anonymized_name"
[51,] "c49individual_name"
[52,] "c50description"
[53,] "c51sample_url_link"
[54,] "c52sample_entrez_link"
[55,] "c53sample_attribute"
[56,] "c54study_ID"
[57,] "c55study_alias"
[58,] "c56study_accession"
[59,] "c57study_title"
[60,] "c58study_type"
[61,] "c59study_abstract"
[62,] "c60center_project_name"
[63,] "c61study_description"
[64,] "c62study_url_link"
[65,] "c63study_entrez_link"
[66,] "c64study_attribute"
[67,] "c65related_studies"
[68,] "c66primary_study"
[69,] "c67submission_ID"
[70,] "c68submission_accession"
[71,] "c69submission_comment"
[72,] "c70submission_center"
[73,] "c71submission_lab"
[74,] "c72submission_date"
[75,] "c73sradb_updated"

table sra_ft_segdir
===================

[,1]
[1,] "level"
[2,] "idx"
[3,] "start_block"
[4,] "leaves_end_block"
[5,] "end_block"
[6,] "root"

table sra_ft_segments
=====================

[,1]
[1,] "blockid"
[2,] "block"

table study
===========

[,1]
 [1,] "study_ID"
 [2,] "study_alias"
 [3,] "study_accession"
 [4,] "study_title"
 [5,] "study_type"
 [6,] "study_abstract"
 [7,] "broker_name"
 [8,] "center_name"
 [9,] "center_project_name"
[10,] "study_description"
[11,] "related_studies"
[12,] "primary_study"
[13,] "sra_link"
[14,] "study_url_link"
[15,] "xref_link"
[16,] "study_entrez_link"
[17,] "ddbj_link"
[18,] "ena_link"
[19,] "study_attribute"
[20,] "submission_accession"
[21,] "sradb_updated"

table submission
================

[,1]
 [1,] "submission_ID"
 [2,] "submission_alias"
 [3,] "submission_accession"
 [4,] "submission_comment"
 [5,] "files"
 [6,] "broker_name"
 [7,] "center_name"
 [8,] "lab_name"
 [9,] "submission_date"
[10,] "sra_link"
[11,] "submission_url_link"
[12,] "xref_link"
[13,] "submission_entrez_link"
[14,] "ddbj_link"
[15,] "ena_link"
[16,] "submission_attribute"
[17,] "sradb_updated"

Il ne reste plus qu'à écrire un petit programme en R pour interroger ces tables de données avec ces numéros, soit les paramètres :

Num Articles                                       Numéros d'accession

1. Bokulich_2014.txt                              (aucun)
 2. Bulgarelli_2012.txt                            SRA043581
 3. Bulgarelli_2015.txt                            PRJEB5860
 4. Horton_2014.txt                                PRJEB7247
 5. Knief_2012.txt                                 HE589809 HE589810 HE589811 HE589812 HE589813 [...]
 6. Lundberg_2012.txt                              ERP001384
 7. Maignien_2014.txt                              SLS_PHY_Bv6v4
 8. Ofek-Lalzar_2014.txt                           PRJNA208116 SAMN02202255 SRX310152 SRX310217 [...]
 9. Ottesen_2013.txt                               SRA061333
 10. Schlaeppi_2014.txt                            PRJEB5058
 11_Telias_2011.txt                                SRX040888
 12_Perazzolli_2014.txt                            PRJNA237062
 13_Furnkranz_2012.txt                             HQ163897 HQ163898 HQ163899 HQ163900 HQ163902 [...]
 14_Carvalhais_2015.txt                            (aucun)
 15_Shi_2015.txt                                   (aucun)
 16_Badri_2013.txt                                 (aucun)
 17_Copeland_2015.txt                              (aucun)
 18_Williams_2013.txt                              SUB148767 SAMNO1882564 SAMNO1882785
 19_Taylor_2014.txt                                SRX091268
 20_Lebeis_2015.txt                                ERP010780 ERP010863
 21_ismej20109a.txt                                (aucun)
 22_journal.pone.0094710.txt                       PRJEB4653
 23_journal.pone.0112763.txt                       SRR1284285 SRR1284286 SRR1284287 SRR1284288 [...]
 24_PNAS-2015-Edwards-E911-20.txt                  SRP044745
 25_ApplEnvironMicrobiol2015-Hernandez-2244-53.txt SRX620410 SRX620411 SRX620412 SRX620413 [...]
 26_MicrobiolMolRev-2006-Casadesus-830-56.txt      X14471 AF109675 S59541 X76688 M98766 [...]

Mais il ne faut pas crier victoire trop vite : les metadata sont souvent mal renseignés voire non renseignés. Voici des exemples de ce qu'on trouve comme renseignements censés nous aider à comprendre ce que contiennent les données

Description des expériences (extrait)
---------------------------

[507] "454 GS FLX Titanium sequencing; P453.exp"
 [509] "454 GS FLX Titanium sequencing; P455.exp"
 [511] "454 GS FLX Titanium sequencing; P457.exp"
 [513] "454 GS FLX Titanium sequencing; P459.exp"
 [515] "454 GS FLX Titanium sequencing; P460.exp"

[1221] "JE.TC.77"
[1223] "JE.TC.79"
[1225] "JE.TC.80"
[1227] "JE.TC.82"
[1229] "JE.TC.9"
[1231] "Ler2_F"
[1233] "R_AtLer15K"
[1235] "R_AtLer18G"
[1237] "R_AtLer21K"
[1239] "R_AtLer26K"
[1241] "R_AtLer34G"
[1243] "R_AtLer36G"
[1245] "R_AtLer4G"

[1267] "Rice Root Micobiome"
[1269] "Rice Root Microbiom"
[1271] "Rice Root Microbiome"
[1273] "root from plants not labeled rep1"
[1275] "root from plants not labeled rep3"
[1277] "Sha2_F"
[1279] "SIP-heavy fraction of rhizosphere from plants labeled with 12CO2 rep2"
[1281] "SIP-heavy fraction of rhizosphere from plants labeled with 13CO2 rep1"
[1283] "SIP-heavy fraction of rhizosphere from plants labeled with 13CO2 rep3"
[1285] "SIP-heavy fraction of root from plants labeled with 12CO2 rep2"
[1287] "SIP-heavy fraction of root from plants labeled with 13CO2 rep1"
[1289] "SIP-heavy fraction of root from plants labeled with 13CO2 rep3"
[1291] "SIP-light fraction of rhizosphere from plants labeled with 12CO2 rep2"
[1293] "SIP-light fraction of rhizosphere from plants labeled with 13CO2 rep1"
[1295] "SIP-light fraction of rhizosphere from plants labeled with 13CO2 rep3"
[1297] "SIP-light fraction of root from plants labeled with 12CO2 rep2"
[1299] "SIP-light fraction of root from plants labeled with 13CO2 rep1"
[1301] "SIP-light fraction of root from plants labeled with 13CO2 rep3"
[1303] "soil whole genome shotgun illumina paired end metagenome samples"
[1304] "Tomato surface and water sources - Multiplex 16S amplicon"
[1305] "Unplanted soil bacterial community"
[1306] "Vine-associated yeasts"
[1307] "wheat rhizosophere illumina paired end whole genome shotgun metagenome samples"

Colonne description, infos triées et uniques (extrait)
======================================================

[1] "454 sequencing plate code: 49a; Sterile 1/2 strength MS media; 1 week old seedlings; host genotype Col-0;"
   [2] "454 sequencing plate code: 49a; Sterile 1/2 strength MS media; 1 week old seedlings; host genotype Cvi-0;"
   [3] "454 sequencing plate code: 49a; Sterile 1/2 strength MS media; 1 week old seedlings; host genotype Shahdara;"
   [4] "454 sequencing plate code: 49a; Sterile 1/2 strength MS media; 1 week old seedlings; host genotype Tsu-0;"
   [5] "Arbuckle_93-11_Endosphere_1"
   [6] "Arbuckle_93-11_Endosphere_2"
   [7] "Arbuckle_93-11_Endosphere_3"
   [8] "Arbuckle_93-11_Endosphere_4"
   [9] "Arbuckle_93-11_Rhizoplane_1"
  [10] "Arbuckle_93-11_Rhizoplane_2"
  [11] "Arbuckle_93-11_Rhizoplane_3"
  [12] "Arbuckle_93-11_Rhizoplane_4"
  [13] "Arbuckle_93-11_Rhizosphere_1"
  [14] "Arbuckle_93-11_Rhizosphere_2"
  [15] "Arbuckle_93-11_Rhizosphere_3"
  [16] "Arbuckle_93-11_Rhizosphere_4"
  [17] "Arbuckle_Glab_B_Endosphere_1"
  [18] "Arbuckle_Glab_B_Endosphere_2"
  [19] "Arbuckle_Glab_B_Endosphere_3"
  [20] "Arbuckle_Glab_B_Endosphere_4"
  [21] "Arbuckle_Glab_B_Rhizoplane_1"
  [22] "Arbuckle_Glab_B_Rhizoplane_2"
  [23] "Arbuckle_Glab_B_Rhizoplane_3"
  [24] "Arbuckle_Glab_B_Rhizoplane_4"
  [25] "Arbuckle_Glab_B_Rhizosphere_1"
  [26] "Arbuckle_Glab_B_Rhizosphere_2"
  [27] "Arbuckle_Glab_B_Rhizosphere_3"
  [28] "Arbuckle_Glab_B_Rhizosphere_4"
  [29] "Arbuckle_Glab_E_Endosphere_1"
  [30] "Arbuckle_Glab_E_Endosphere_2"
  [31] "Arbuckle_Glab_E_Endosphere_3"
  [32] "Arbuckle_Glab_E_Endosphere_4"
  [33] "Arbuckle_Glab_E_Rhizoplane_1"
  [34] "Arbuckle_Glab_E_Rhizoplane_2"
  [35] "Arbuckle_Glab_E_Rhizoplane_3"
  [36] "Arbuckle_Glab_E_Rhizoplane_4"
  [37] "Arbuckle_Glab_E_Rhizosphere_1"

[219] "Endophyte compartment collected from within surface-cleaned roots; 454 sequencing plate code: 29a;
        young developmental stage (first appearance of inflorescence); host genotype Tsu-0; Mason Farm replicate 1;"

[220] "Endophyte compartment collected from within surface-cleaned roots; 454 sequencing plate code: 35a;
        old developmental stage (~4 weeks after yng stage with 50% of leaves sensencent); host genotype Col-0; Mason Farm replicate 1;"

[221] "Endophyte compartment collected from within surface-cleaned roots; 454 sequencing plate code: 35a;
        old developmental stage (~4 weeks after yng stage with 50% of leaves sensencent); host genotype Ct-0; M

De plus il faut trouver la taille des données et non pas tout télécharger. Heureusement, nous avons finalement réussi à trouver une option du logiciel wget qui donne en ligne de commandes la taille d'un fichier dont on passe l'URI, comme par exemple ici la taille 12688427 -- soit environ 13 Mo -- pour le fichier SRR1524567.sra.

Pour les autres bases, pas de package, pas de fonction et donc soit on attend un hypothétique package R soit il faudra écrire un programme R pour interfacer les base des données... Ainsi, pour le numéro ERP001384, il est possible de consulter les informations en ligne dans un navigateur, à l'aide du lien ci-dessous

Ceci s'obtient facilement si on tape ce numéro d'accession dans le panneau Search de l'ENA

Mais entre parcourir des informations dans une page Web et construire un fichier de résultats, il y a un un grand travail à fournir...

Au final, de façon sûre et automatique, voici ce que nous avons obtenu pour 26 articles sélectionnés

Article                                             Source  Fichiers   Taille
1. Bokulich_2014.txt                                ENA ?          0        0
2. Bulgarelli_2012.txt                              SRA          130     1020 Mo
3. Bulgarelli_2015.txt                              SRA            0        0
4. Horton_2014.txt                                  SRA            0        0
5. Knief_2012.txt                                   SRA            0        0
6. Lundberg_2012.txt                                SRA         1277       10 Go
7. Maignien_2014.txt                                SRA            0        0
8. Ofek-Lalzar_2014.txt                             SRA           10       33 Go
9. Ottesen_2013.txt                                 SRA           54        2 Go
10. Schlaeppi_2014.txt                              SRA            0        0
11_Telias_2011.txt                                  SRA            1       30 Mo
12_Perazzolli_2014.txt                              SRA            0        0
13_Furnkranz_2012.txt                               SRA            0        0
14_Carvalhais_2015.txt                              ENA ?          0        0
15_Shi_2015.txt                                     ENA ?          0        0
16_Badri_2013.txt                                   ENA ?          0        0
17_Copeland_2015.txt                                ENA ?          0        0
18_Williams_2013.txt                                SRA            0        0
19_Taylor_2014.txt                                  SRA            9      156 Mo
20_Lebeis_2015.txt                                  SRA          541        7 Go
21_ismej20109a.txt                                  ENA ?          0        0
22_journal.pone.0094710.txt                         SRA            0        0
23_journal.pone.0112763.txt                         SRA            1       14 Mo
24_PNAS-2015-Edwards-E911-20.txt                    SRA          482        6 Go
25_ApplEnvironMicrobiol2015-Hernandez-2244-53.txt   SRA            1       11 Mo
26_MicrobiolMolRev-2006-Casadesus-830-56.txt        SRA            0        0

Et voici ce que nous avons "bricolé" en semi-automatique avec une intervention humaine :

Il est clair que ce n'est pas très satisfaisant pour décider si ces données sont pertinentes pour nos études dans l'axe 4.

Conseil Scientifique METASEED : axe 4

(mardi 8 décembre 2015)

Table des matières cliquable

1. Rappel de l'axe 4

2. Collecte des données

3. Automatisation de l'exploitation

4. Bilan