update memory requirements

Ecogenomics · Apr 22, 2021 · 61e1e88 · 61e1e88
1 parent 34d9474
commit 61e1e88
Show file tree

Hide file tree

Showing 4 changed files with 24 additions and 22 deletions.
diff --git a/docs/src/installing/index.rst b/docs/src/installing/index.rst
@@ -34,11 +34,11 @@ Hardware requirements
      - Storage
      - Time
    * - Archaea
-     - ~8 GB
+     - ~13 GB
      - ~27 GB
      - ~1 hour / 1,000 genomes @ 64 CPUs
    * - Bacteria
-     - ~150 GB
+     - ~204 GB
      - ~27 GB
      - ~1 hour / 1,000 genomes @ 64 CPUs
 

diff --git a/gtdbtk/classify.py b/gtdbtk/classify.py
@@ -132,21 +132,21 @@ def place_genomes(self,
         """Place genomes into reference tree using pplacer."""
 
         # Warn if the memory is insufficient
+        mem_warning = 'pplacer requires ~{req_gb} GB of RAM to fully load the ' \
+                      '{domain} tree into memory. However, {cur_gb:,} GB was ' \
+                      'detected. This may affect pplacer performance, or fail' \
+                      ' if there is insufficient swap space.'
         mem_gb = get_memory_gb()
         if mem_gb is not None:
             mem_total = mem_gb['MemTotal']
-            if marker_set_id == 'bac120' and mem_total < 145:
-                self.logger.warning(f'pplacer requires ~152 GB of RAM to fully '
-                                    f'load the bacterial tree into memory. '
-                                    f'However, {mem_total:,}GB was detected. '
-                                    f'This may affect pplacer performance, '
-                                    f'or fail if there is insufficient scratch space.')
-            elif marker_set_id == 'ar122' and mem_total < 6:
-                self.logger.warning(f'pplacer requires ~8.2 GB of RAM to fully '
-                                    f'load the archaeal tree into memory. '
-                                    f'However, {mem_total:,}GB was detected. '
-                                    f'This may affect pplacer performance, '
-                                    f'or fail if there is insufficient scratch space.')
+            if marker_set_id == 'bac120' and mem_total < Config.PPLACER_MIN_RAM_BAC:
+                self.logger.warning(mem_warning.format(req_gb=Config.PPLACER_MIN_RAM_BAC,
+                                                       domain='bacterial',
+                                                       cur_gb=mem_total))
+            elif marker_set_id == 'ar122' and mem_total < Config.PPLACER_MIN_RAM_ARC:
+                self.logger.warning(mem_warning.format(req_gb=Config.PPLACER_MIN_RAM_ARC,
+                                                       domain='archaeal',
+                                                       cur_gb=mem_total))
 
         # rename user MSA file for compatibility with pplacer
         if not user_msa_file.endswith('.fasta'):

diff --git a/gtdbtk/config/config.py b/gtdbtk/config/config.py
@@ -196,6 +196,8 @@
 PPLACER_BAC120_REF_PKG = f"gtdb_{VERSION_DATA}_bac120.refpkg"
 PPLACER_AR122_REF_PKG = f"gtdb_{VERSION_DATA}_ar122.refpkg"
 PPLACER_RPS23_REF_PKG = f"gtdb_{VERSION_DATA}_rps23.refpkg"
+PPLACER_MIN_RAM_BAC = 204
+PPLACER_MIN_RAM_ARC = 13
 
 # Fastani configuration
 FASTANI_SPECIES_THRESHOLD = 95.0
@@ -208,13 +210,13 @@
 MRCA_RED_AR122 = os.path.join(RED_DIR, f"gtdbtk_{VERSION_DATA}_ar122.tsv")
 
 # Hashing information for validating the reference package.
-REF_HASHES = {PPLACER_DIR: 'f41cfe0284ebaca4485b42e054936190c6a88bd1',
-              MASK_DIR: '63551a43333bc6cbc9abf139ce881847ca19240b',
-              MARKER_DIR: 'a325720422d8348d7a934143cc86112b6c92ac98',
-              RADII_DIR: '1092727925f38a8a2b3f4fb40e3316c0083671f5',
-              MSA_FOLDER: 'cf91d712c733e7e2535a41e6153c12b3c37d1ede',
-              METADATA_DIR: 'e003b4d5d48302e85c536751f663a70447de83d4',
-              TAX_FOLDER: '30c5970b2eaf5df654b2e01bfa39265302c0be89',
+REF_HASHES = {PPLACER_DIR: '4d931b5109a240602f55228029b87ee768da8141',
+              MASK_DIR: '36d6ac371d247b2b952523b9798e78908ea323fa',
+              MARKER_DIR: '2ba5ae35fb272462663651d18fd9e523317e48cd',
+              RADII_DIR: '9f9a2e21e27b9049044d04d731795499414a365c',
+              MSA_FOLDER: 'b426865245c39ee9f01b0392fb8f7867a9f76f0a',
+              METADATA_DIR: '7640aed96fdb13707a2b79b746a94335faabd6df',
+              TAX_FOLDER: '4a7a1e4047c088e92dee9740206499cdb7e5beca',
               FASTANI_DIR: '6a3555bb61d9cc3163c26e65772b96b8f58a2d84',
               RED_DIR: '6f661eef8e172a8a7e78af2a74fe4d079a3f5b0f'}
 

diff --git a/gtdbtk/misc.py b/gtdbtk/misc.py
@@ -158,7 +158,7 @@ def check_install(self):
 
             if user_hash != expected_hash:
                 self.logger.info("         |-- {:16} {}".format(
-                    base_name, colour('HASH MISMATCH', ['bright'], fg='yellow')))
+                    base_name, colour(f'HASH MISMATCH {user_hash}', ['bright'], fg='yellow')))
                 ok = False
             else:
                 self.logger.info("         |-- {:16} {}".format(