zipstrain 0.2.8__tar.gz → 0.2.11__tar.gz

{zipstrain-0.2.8 → zipstrain-0.2.11}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: zipstrain
-Version: 0.2.8
+Version: 0.2.11
 Summary: 
 Author: ParsaGhadermazi
 Author-email: 54489047+ParsaGhadermazi@users.noreply.github.com

{zipstrain-0.2.8 → zipstrain-0.2.11}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "zipstrain"
-version = "0.2.8"
+version = "0.2.11"
 description = ""
 authors = [
     {name = "ParsaGhadermazi",email = "54489047+ParsaGhadermazi@users.noreply.github.com"}

{zipstrain-0.2.8 → zipstrain-0.2.11}/src/zipstrain/cli.py

@@ -324,43 +324,91 @@ def to_complete_table(genome_comparison_object, output_file):
 @click.option('--profile-file', '-p', required=True, help="Path to the profile Parquet file.")
 @click.option('--stb-file', '-s', required=True, help="Path to the scaffold-to-genome mapping file.")
 @click.option('--bed-file', '-b', required=True, help="Path to the BED file.")
-@click.option('--min-cov', '-c', default=0.1, help="Minimum genome-wide coverage to use homogeneous Poisson point process.")
-@click.option('--ber', '-e', default=0.5, help="Minimum breadth to expected breadth ratio to consider a genome as present.")
-@click.option('--cv-threshold', '-v', default=1.0, help="Maximum coefficient of variation to consider genome as present when coverage is smaller than min-cov.")
+@click.option('--read-loc-file', '-r', required=True, help="Path to the read location table.")
+@click.option('--min-cov-fug', '-c', default=0.1, help="Minimum coverage to use fug.")
+@click.option('--fug-threshold', '-f', default=2, help="FUG threshold.")
+@click.option('--ber', '-e', default=0.5, help="Minimum ratio of breadth over expected breadth to consider presence.")
 @click.option('--output-file', '-o', required=True, help="Path to save the output Parquet file.")
-def presence_profile(profile_file, stb_file, bed_file, min_cov, ber, cv_threshold, output_file):
+def presence_profile(profile_file, stb_file, bed_file, read_loc_file, min_cov_fug, fug_threshold, ber, output_file):
     """
-    Generate a presence profile from the given files.
+    Generate a presence profile for genomes based on the given profile and read location data.
 
     Args:
     profile_file (str): Path to the profile Parquet file.
     stb_file (str): Path to the scaffold-to-genome mapping file.
     bed_file (str): Path to the BED file.
+    read_loc_file (str): Path to the read location table.
+    min_cov_fug (float): Minimum coverage to use fug.
+    fug_threshold (float): FUG threshold.
+    ber (float): Minimum ratio of breadth over expected breadth to consider presence.
+    output_file (str): Path to save the output Parquet file.
     """
-    profile=pl.scan_parquet(profile_file)
+    profile = pl.scan_parquet(profile_file)
     stb = pl.scan_csv(stb_file, separator="\t", has_header=False).with_columns(
         pl.col("column_1").alias("scaffold"),
         pl.col("column_2").alias("genome")
     ).select(["scaffold", "genome"])
     bed = pl.scan_csv(bed_file, separator="\t", has_header=False).with_columns(
         pl.col("column_1").alias("scaffold"),
-        pl.col("column_2").alias("start"),
-        pl.col("column_3").alias("end")
+        pl.col("column_2").cast(pl.Int64).alias("start"),
+        pl.col("column_3").cast(pl.Int64).alias("end")
     ).select(["scaffold", "start", "end"])
-    ut.estimate_genome_presence(
+    read_loc_table = pl.scan_parquet(read_loc_file).rename({
+        "chrom":"scaffold",
+        "pos":"loc"
+    })
+    presence_df = ut.get_genome_stats(
         profile=profile,
         stb=stb,
         bed=bed,
-        cv_threshold=cv_threshold,
-        ber=ber,
-        min_cov_constant_poisson=min_cov
-    ).sink_parquet(output_file, compression='zstd',engine="streaming")
+        read_loc_table=read_loc_table,
+        min_cov_use_fug=min_cov_fug,
+        fug=fug_threshold,
+        ber=ber
+    )
+    presence_df.sink_parquet(output_file, compression='zstd')
+
+@utilities.command("process-read-locs")
+@click.option("--output-file", "-o", required=True, help="Path to save the processed read locations Parquet file.")
+def process_read_locs(output_file):
+    """
+    Process read locations and save them to a Parquet file.
+
+    Args:
+    output_file (str): Path to save the output Parquet file.
+    """
+    ut.process_read_location(output_file=pathlib.Path(output_file))
 
 @cli.group()
 def gene_tools():
     """Holds anything related to gene analysis."""
     pass
 
+@utilities.command("generate_stb")
+@click.option('--genomes-dir-file', '-g', required=True, help="Path to the genomes directory file. A text file with each line containing a genome fasta file path.")
+@click.option('--output-file', '-o', required=True, help="Path to save the output scaffold-to-genome mapping file.")
+@click.option('--extension', '-e', default=".fasta", help="File extension of the genome fasta files.")
+def generate_stb(genomes_dir_file, output_file, extension):
+    """
+    Generate a scaffold-to-genome mapping file from the given genomes directory file.
+
+    Args:
+    genomes_dir_file (str): Path to the genomes directory file.
+    output_file (str): Path to save the output scaffold-to-genome mapping file.
+    extension (str): File extension of the genome fasta files.
+    """
+    with open(output_file, 'w') as out_f:
+        for genome in pathlib.Path(genomes_dir_file).glob(f"*{extension}"):
+            genome_name = genome.stem
+            with open(genome, 'r') as gf:
+                for line in gf:
+                    if line.startswith('>'):
+                        scaffold_name = line[1:].strip().split()[0]
+                        out_f.write(f"{scaffold_name}\t{genome_name}\n")
+    
+        
+    
+
 
 @gene_tools.command("gene-range-table")
 @click.option('--gene-file', '-g', required=True, help="location of gene file. Prodigal's nucleotide fasta output")
@@ -603,22 +651,34 @@ def prepare_profiling(reference_fasta, gene_fasta, stb_file, output_dir):
 @profile.command("profile-single")
 @click.option('--bed-file', '-b', required=True, help="Path to the BED file describing regions to be profiled.")
 @click.option('--bam-file', '-a', required=True, help="Path to the BAM file to be profiled.")
+@click.option('--stb-file', '-s', required=True, help="Path to the scaffold-to-genome mapping file.")
 @click.option('--gene-range-table', '-g', required=True, help="Path to the gene range table.")
 @click.option('--num-workers', '-n', default=1, help="Number of workers to use for profiling.")
 @click.option('--output-dir', '-o', required=True, help="Directory to save the profiling output.")
-def profile_single(bed_file, bam_file, gene_range_table, num_workers, output_dir):
+@click.option('--ber', '-r', default=0.5, help="Minimum ratio of breadth over expected breadth to consider presence.")
+@click.option('--fug', '-f', default=2.0, help="fraction of expected gaps (FUG) threshold.")
+@click.option('--min-cov-use-fug', '-m', default=0.1, help="Minimum coverage to use FUG.")
+def profile_single(bed_file, bam_file, stb_file, gene_range_table, num_workers, output_dir, ber, fug, min_cov_use_fug):
     """
     Profile a single BAM file using the provided BED file and gene range table.
     
     """
     output_dir=pathlib.Path(output_dir)
     output_dir.mkdir(parents=True, exist_ok=True)
+    stb= pl.scan_csv(stb_file, separator='\t',has_header=False).with_columns(
+        pl.col("column_1").alias("scaffold"),
+        pl.col("column_2").alias("genome")
+    )
     pf.profile_bam(
         bed_file=bed_file,
         bam_file=bam_file,
         gene_range_table=gene_range_table,
+        stb=stb,
         output_dir=output_dir,
-        num_workers=num_workers
+        num_workers=num_workers,
+        ber=ber,
+        fug=fug,
+        min_cov_use_fug=min_cov_use_fug
     )
 
 @cli.group()

{zipstrain-0.2.8 → zipstrain-0.2.11}/src/zipstrain/profile.py

@@ -147,13 +147,27 @@ async def _profile_chunk_task(
     stdout, stderr = await proc.communicate()
     if proc.returncode != 0:
         raise Exception(f"Command failed with error: {stderr.decode().strip()}")
+    cmd=["samtools", "view", "-F", "132", "-L", str(bed_file.absolute()), str(bam_file.absolute()), "|", "zipstrain", "utilities", "process-read-locs", "--output-file", f"{bam_file.stem}_read_locs_{chunk_id}.parquet"]
+    proc = await asyncio.create_subprocess_shell(
+                " ".join(cmd),
+                stdout=asyncio.subprocess.PIPE,
+                stderr=asyncio.subprocess.PIPE,
+                cwd=output_dir
+            )
+    stdout, stderr = await proc.communicate()
+    if proc.returncode != 0:
+        raise Exception(f"Command failed with error: {stderr.decode().strip()}")
 
 async def profile_bam_in_chunks(
     bed_file:str,
     bam_file:str,
     gene_range_table:str,
+    stb:pl.LazyFrame,
     output_dir:str,
-    num_workers:int=4
+    num_workers:int=4,
+    ber:float=0.5,
+    fug:float=2.0,
+    min_cov_use_fug:int=0.1
 )->None:
     """
     Profile a BAM file in chunks using provided BED files.
@@ -189,17 +203,46 @@ async def profile_bam_in_chunks(
             chunk_id=chunk_id
         ))
     await asyncio.gather(*tasks) 
-    pfs=[output_dir/"tmp"/f"{bam_file.stem}_{chunk_id}.parquet" for chunk_id in range(len(bed_chunk_files)) if (output_dir/"tmp"/f"{bam_file.stem}_{chunk_id}.parquet").exists()]
-    mpileup_df = pl.concat([pl.scan_parquet(pf) for pf in pfs])
-    mpileup_df.sink_parquet(output_dir/f"{bam_file.stem}.parquet", compression='zstd')
+    pfs=[(output_dir/"tmp"/f"{bam_file.stem}_{chunk_id}.parquet", output_dir/"tmp"/f"{bam_file.stem}_read_locs_{chunk_id}.parquet" ) for chunk_id in range(len(bed_chunk_files)) if (output_dir/"tmp"/f"{bam_file.stem}_{chunk_id}.parquet").exists()]
+
+    mpile_container: list[pl.LazyFrame] = []
+    read_loc_pfs: list[pl.LazyFrame] = []
+    for pf, read_loc_pf in pfs:
+        mpile_container.append(pl.scan_parquet(pf).lazy())
+        read_loc_pfs.append(pl.scan_parquet(read_loc_pf).lazy())
+
+    mpileup_df = pl.concat(mpile_container)
+    mpileup_df.sink_parquet(output_dir/f"{bam_file.stem}_profile.parquet", compression='zstd', engine='streaming')
+    read_loc_df = pl.concat(read_loc_pfs).rename(
+        {
+            "chrom":"scaffold",
+            "pos":"loc",
+        }
+    )
+
+    utils.get_genome_stats(
+        profile=mpileup_df,
+        read_loc_table=read_loc_df,
+        stb=stb,
+        bed=bed_lf.rename({"column_1":"scaffold","column_2":"start","column_3":"end"}),
+        ber=ber,
+        fug=fug,
+        min_cov_use_fug=min_cov_use_fug,
+    ).sink_parquet(output_dir/f"{bam_file.stem}_genome_stats.parquet", compression='zstd', engine='streaming')
+    
+    
     os.system(f"rm -r {output_dir}/tmp")
 
 def profile_bam(
     bed_file:str,
     bam_file:str,
     gene_range_table:str,
+    stb:pl.LazyFrame,
     output_dir:str,
-    num_workers:int=4
+    num_workers:int=4,
+    ber:float=0.5,
+    fug:float=2.0,
+    min_cov_use_fug:int=0.1
 )->None:
     """
     Profile a BAM file in chunks using provided BED files.
@@ -208,6 +251,7 @@ def profile_bam(
     bed_file (list[pathlib.Path]): A bed file describing all regions to be profiled.
     bam_file (pathlib.Path): Path to the BAM file.
     gene_range_table (pathlib.Path): Path to the gene range table.
+    stb (pl.LazyFrame): Scaffold-to-bin mapping table.
     output_dir (pathlib.Path): Directory to save output files.
     num_workers (int): Number of concurrent workers to use.
     """
@@ -215,7 +259,11 @@ def profile_bam(
         bed_file=bed_file,
         bam_file=bam_file,
         gene_range_table=gene_range_table,
+        stb=stb,
         output_dir=output_dir,
-        num_workers=num_workers
+        num_workers=num_workers,
+        ber=ber,
+        fug=fug,
+        min_cov_use_fug=min_cov_use_fug
     ))
 

{zipstrain-0.2.8 → zipstrain-0.2.11}/src/zipstrain/task_manager.py

@@ -127,6 +127,9 @@ class Status(StrEnum):
     SUCCESS = "success"
     PENDING = "pending"
 
+class Messages(StrEnum):
+    """Enumeration of common messages used in task and batch management."""
+    CANCELLED_BY_USER = "Task was cancelled by a signal from the user."
 
 class Input(ABC):
     """Abstract base class for task inputs. DO NOT INSTANTIATE DIRECTLY.
@@ -275,7 +278,6 @@ class IntOutput(Output):
             raise ValueError(f"Output value for task {self.task.id} is not an integer.")
         else:
             return False
-        return False
 
 
 class Engine(ABC):
@@ -391,6 +393,7 @@ class Task(ABC):
         """Asynchronously reads the task status from the .status file in the task directory."""
         status_path = self.task_dir / ".status"
         # read the status file if it exists
+        
         if status_path.exists():
             raw = await read_file(status_path, self.file_semaphore)
             self._status = raw.strip()
@@ -406,12 +409,11 @@ class Task(ABC):
                 except Exception:
                     all_ready = False
 
-                if all_ready:
+                if all_ready or self._batch_obj._cleaned_up:
                     self._status = Status.SUCCESS.value
-                    await write_file(status_path, Status.SUCCESS.value, self.file_semaphore)
+                    
                 else:
                     self._status = Status.FAILED.value
-                    await write_file(status_path, Status.FAILED.value, self.file_semaphore)
                     raise ValueError(f"Task {self.id} reported done but outputs are not ready or invalid. {self.expected_outputs['output-file'].expected_file.absolute()}")
 
         return self._status
@@ -526,8 +528,8 @@ class ProfileTaskGenerator(TaskGenerator):
                 }
                 expected_outputs ={
                 "profile":  FileOutput(row["sample_name"]+".parquet" ),
-                "breadth":  FileOutput(row["sample_name"]+"_breadth.parquet" ),
                 "scaffold": FileOutput(row["sample_name"]+".parquet.scaffolds" ),
+                "genome-stats": FileOutput(row["sample_name"]+"_genome_stats.parquet" ),
                 }
                 task = ProfileBamTask(id=row["sample_name"], inputs=inputs, expected_outputs=expected_outputs, engine=self.engine)
                 tasks.append(task)
@@ -637,9 +639,8 @@ class Batch(ABC):
             task.map_io()
         
         self._runner_obj:Runner = None
-            
+        self._cleaned_up = False
 
-    
     def _get_initial_status(self) -> str:
         """Returns the initial status of the batch based on the presence of the batch directory."""
         if not self.batch_dir.exists():
@@ -663,6 +664,7 @@ class Batch(ABC):
 
     def cleanup(self) -> None:
         """The base class defines if any cleanup is needed after batch success. By default, it does nothing."""
+        self._cleaned_up = True
         return None
 
     @abstractmethod
@@ -724,27 +726,24 @@ class LocalBatch(Batch):
 
     async def run(self) -> None:
         """This method runs all tasks in the batch locally by creating a shell script and executing it."""
-        if self.status != Status.SUCCESS and self.status != Status.FAILED.value:
+        if self.status != Status.SUCCESS:
             self.batch_dir.mkdir(parents=True, exist_ok=True)
+            
             self._status = Status.RUNNING.value
-
-
             await write_file(self.batch_dir / ".status", self._status, self.file_semaphore)
-
+            
+            script_path = self.batch_dir / f"{self.id}.sh" # Path to the shell script for the batch
+            script = self._script # Initialize the script content
+            
             for task in self.tasks:
-                if task.status == Status.NOT_STARTED.value:
+                if task.status != Status.SUCCESS.value:
                     task.task_dir.mkdir(parents=True, exist_ok=True)  # Create task directory
                     await write_file(task.task_dir / ".status", Status.NOT_STARTED.value, self.file_semaphore)
-
-            script_path = self.batch_dir / f"{self.id}.sh" # Path to the shell script for the batch
-
-            script = self._script
-            for task in self.tasks:
-                if task.status == Status.NOT_STARTED.value or task.status == Status.FAILED.value:
                     script += f"\n{task.pre_run}\n{task.command}\n{task.post_run}\n"
             
+            
+            
             await write_file(script_path, script, self.file_semaphore)
-
             await write_file(self.batch_dir / ".status", self._status, self.file_semaphore)
 
             self._proc = await asyncio.create_subprocess_exec(
@@ -753,12 +752,17 @@ class LocalBatch(Batch):
                 stderr=asyncio.subprocess.PIPE,
                 cwd=self.batch_dir,
             )
+
             try:
                 out_bytes, err_bytes = await self._proc.communicate()
+            
             except asyncio.CancelledError:
                 if self._proc and self._proc.returncode is None:
                     self._proc.terminate()
+                    await write_file(self.batch_dir / f"{self.id}.err", Messages.CANCELLED_BY_USER, self.file_semaphore)
+
                 raise
+            
 
             await write_file(self.batch_dir / f"{self.id}.out", out_bytes.decode(), self.file_semaphore)
             await write_file(self.batch_dir / f"{self.id}.err", err_bytes.decode(), self.file_semaphore)
@@ -767,14 +771,14 @@ class LocalBatch(Batch):
                 self.cleanup()
                 self._status = Status.SUCCESS.value
                 await write_file(self.batch_dir / ".status", self._status, self.file_semaphore)
+            
             else:
                 self._status = Status.FAILED.value
                 await write_file(self.batch_dir / ".status", self._status, self.file_semaphore)
-        
-        elif self.status == Status.SUCCESS.value and self.outputs_ready():
-            self._status = Status.SUCCESS.value
+
         else:
-            self._status = Status.FAILED.value
+            self._status = Status.SUCCESS.value
+
     
     def _parse_job_id(self, sbatch_output):
         return super()._parse_job_id(sbatch_output)
@@ -1470,15 +1474,11 @@ class ProfileBamTask(Task):
     zipstrain profile profile-single --bam-file input.bam \
     --bed-file bed_file.bed \
     --gene-range-table gene-range-table.bed \
+    --stb-file <stb-file> \
     --num-workers <num-workers> \
     --output-dir .
     mv input.bam.parquet <sample-name>.parquet
     samtools idxstats <bam-file> |  awk '$3 > 0 {print $1}' > <sample-name>.parquet.scaffolds
-    zipstrain utilities genome_breadth_matrix --profile <sample-name>.parquet \
-        --genome-length <genome-length-file> \
-        --stb <stb-file> \
-        --min-cov <breadth-min-cov> \
-        --output-file <sample-name>_breadth.parquet
     """
     
 class FastCompareTask(Task):
@@ -1552,6 +1552,7 @@ class FastCompareLocalBatch(LocalBatch):
         for task in tasks_to_remove:
             self.tasks.remove(task)
             shutil.rmtree(task.task_dir)
+        self._cleaned_up = True
 
 class FastCompareSlurmBatch(SlurmBatch):
     """A SlurmBatch that runs FastCompareTask tasks on a Slurm cluster. Maybe removed in future"""
@@ -1560,6 +1561,8 @@ class FastCompareSlurmBatch(SlurmBatch):
         for task in tasks_to_remove:
             self.tasks.remove(task)
             shutil.rmtree(task.task_dir)
+        
+        self._cleaned_up = True
 
 class PrepareCompareGenomeRunOutputsLocalBatch(LocalBatch):
     pass
@@ -1925,8 +1928,11 @@ class FastGeneCompareLocalBatch(LocalBatch):
     def cleanup(self) -> None:
         tasks_to_remove = [task for task in self.tasks if isinstance(task, FastGeneCompareTask)]
         for task in tasks_to_remove:
+            task._status=Status.SUCCESS
             self.tasks.remove(task)
             shutil.rmtree(task.task_dir)
+        self._cleaned_up = True
+        
 
 class FastGeneCompareSlurmBatch(SlurmBatch):
     """A SlurmBatch that runs FastGeneCompareTask tasks on a Slurm cluster."""
@@ -1935,6 +1941,7 @@ class FastGeneCompareSlurmBatch(SlurmBatch):
         for task in tasks_to_remove:
             self.tasks.remove(task)
             shutil.rmtree(task.task_dir)
+        self._cleaned_up = True
 
 class PrepareGeneCompareRunOutputsLocalBatch(LocalBatch):
     pass

{zipstrain-0.2.8 → zipstrain-0.2.11}/src/zipstrain/utils.py

@@ -140,7 +140,7 @@ def process_mpileup_function(gene_range_table_loc, batch_bed, batch_size, output
 
         if writer is None:
             # Open writer for the first time
-            writer = pq.ParquetWriter(output_file, schema, compression='snappy')
+            writer = pq.ParquetWriter(output_file, schema, compression='zstd')
         writer.write_table(pa.Table.from_batches([batch]))
 
         # Clear buffers
@@ -180,6 +180,51 @@ def process_mpileup_function(gene_range_table_loc, batch_bed, batch_size, output
     if writer:
         writer.close()
 
+def process_read_location(output_file:str, batch_size:int=10000)->None:
+    """
+    This function takes the output of samtools view -F 132 and processes it to extract read locations in a parquet file.
+    """
+    schema = pa.schema([
+        ('chrom', pa.string()),
+        ('pos', pa.int32()),
+    ])
+    writer = None
+    chroms = []
+    positions = []
+    def flush_batch():
+        nonlocal writer
+        if not chroms:
+            return
+        batch = pa.RecordBatch.from_arrays([
+            pa.array(chroms, type=pa.string()),
+            pa.array(positions, type=pa.int32()),
+        ], schema=schema)
+
+        if writer is None:
+            # Open writer for the first time
+            writer = pq.ParquetWriter(output_file, schema, compression='zstd')
+        writer.write_table(pa.Table.from_batches([batch]))
+
+        # Clear buffers
+        chroms.clear()
+        positions.clear()
+    for line in sys.stdin:
+        if not line.strip():
+            continue
+        fields = line.strip().split('\t')
+        if len(fields) < 4:
+            continue
+        chrom, pos = fields[2], fields[3]
+        chroms.append(chrom)
+        positions.append(int(pos))
+        if len(chroms) >= batch_size:
+            flush_batch()
+    # Flush remaining data
+    flush_batch()
+    if writer:
+        writer.close()
+
+
 def extract_genome_length(stb: pl.LazyFrame, bed_table: pl.LazyFrame) -> pl.LazyFrame:
     """
     Extract the genome length information from the scaffold-to-genome mapping table.
@@ -350,102 +395,106 @@ def split_lf_to_chunks(lf:pl.LazyFrame,num_chunks:int)->list[pl.LazyFrame]:
     return chunks
 
 
-def estimate_genome_presence(
+def get_genome_gaps(
+    read_loc_table: pl.LazyFrame,
+    stb: pl.LazyFrame,
+    genome_length: pl.LazyFrame,
+                    )-> pl.LazyFrame:
+    read_loc_table=read_loc_table.sort(["scaffold",'loc'])
+    read_loc_table=read_loc_table.with_columns(
+        (pl.col("loc") - pl.col("loc").shift(1).over("scaffold")).alias("gap_length")
+    ).join(
+        stb,
+        on="scaffold",
+        how="left"
+    )
+    delta=read_loc_table.group_by("genome").agg(
+        rn=pl.len()).join(
+        genome_length,
+        on="genome",
+        how="left"
+    ).with_columns(
+        delta=(pl.col("genome_length")/pl.col("rn")).round().alias("delta")).select(
+        pl.col("genome"),
+        pl.col("delta"),
+        pl.col("rn")
+        )
+    read_loc_table=read_loc_table.join(
+        delta,
+        on="genome",
+        how="left"
+    )
+    read_loc_table=read_loc_table.filter(
+        pl.col("gap_length") > pl.col("delta")
+    ).group_by(["genome","gap_length"]).agg(
+        pd=(pl.len()/(pl.col("rn").first()-1)),
+        delta=pl.col("delta").first()
+    ).with_columns(
+        pd= pl.col("pd") * (pl.col("gap_length")-pl.col("delta"))
+    ).group_by("genome").agg(
+        fug=(pl.col("delta").first()-pl.col("pd").sum())/pl.col("delta").first()
+    )
+    return read_loc_table.select(
+        pl.col("genome"),
+        pl.col("fug")
+    )
+
+def get_genome_stats(
     profile:pl.LazyFrame,
     bed: pl.LazyFrame,
     stb: pl.LazyFrame,
+    read_loc_table: pl.LazyFrame,
     ber:float=0.5,
-    cv_threshold:float=2.5,
-    min_cov_constant_poisson: int = 0.5,
+    fug:float=2,
+    min_cov_use_fug:int=0.1
 )->pl.LazyFrame:
-    """
-    This function estimates the presence of genomes in a sample based on coverage information.
-    as long as the coverage is above a certain threshold. BER is used to decide the threshold.
-    However, if the coverage is below the threshold, the coefficient of variation (CV) is used instead as
-    a more reliable metric for low-coverage scenarios.
-    
-    Args:
-        profile (pl.LazyFrame): The profile LazyFrame containing coverage information.
-        bed (pl.LazyFrame): The BED table containing genomic regions.
-        stb (pl.LazyFrame): The scaffold-to-bin mapping LazyFrame.
-        ber (float): Breadth over expected breadth ratio threshold for genome presence.
-        cv_threshold (float): Coefficient of variation threshold for genome presence.
-        min_cov_constant_poisson (int): Minimum coverage threshold to use BER for presence estimation.
-        
-    Returns:
-        pl.LazyFrame: A LazyFrame containing genome presence information.
-    """
-    profile=profile.with_columns(
-        (pl.col("A")+pl.col("T")+pl.col("C")+pl.col("G")).alias("coverage")
-        )
-    starts_df=bed.select(
-        pl.col("scaffold").cast(profile.collect_schema()["chrom"]).alias("chrom"),
-        pl.col("start").cast(profile.collect_schema()["pos"]).alias("pos"),
-        pl.lit("NA").cast(profile.collect_schema()["gene"]).alias("gene"),
-        pl.lit(0).cast(profile.collect_schema()["A"]).alias("A"),
-        pl.lit(0).cast(profile.collect_schema()["T"]).alias("T"),
-        pl.lit(0).cast(profile.collect_schema()["C"]).alias("C"),
-        pl.lit(0).cast(profile.collect_schema()["G"]).alias("G"),
-        pl.lit(0).cast(profile.collect_schema()["coverage"]).alias("coverage")
-        )
-    ends_df=bed.select(
-        pl.col("scaffold").cast(profile.collect_schema()["chrom"]).alias("chrom"),
-        (pl.col("end")-1).cast(profile.collect_schema()["pos"]).alias("pos"),
-        pl.lit("NA").cast(profile.collect_schema()["gene"]).alias("gene"),
-        pl.lit(0).cast(profile.collect_schema()["A"]).alias("A"),
-        pl.lit(0).cast(profile.collect_schema()["T"]).alias("T"),
-        pl.lit(0).cast(profile.collect_schema()["C"]).alias("C"),
-        pl.lit(0).cast(profile.collect_schema()["G"]).alias("G"),
-        pl.lit(0).cast(profile.collect_schema()["coverage"]).alias("coverage")
-        )
 
-    profile=pl.concat([profile,starts_df,ends_df]).unique(subset=["chrom","pos"],keep="first").sort(["chrom","pos"])
-    genome_lengths=bed.join(
+    genome_lengths=extract_genome_length(stb, bed)
+    genome_gap_stats= get_genome_gaps(read_loc_table, stb, genome_lengths)
+    profile=profile.join(
         stb,
-        on="scaffold",
+        left_on="chrom",
+        right_on="scaffold",
         how="left"
-    ).group_by("genome").agg(
-        genome_length=(pl.col("end") - pl.col("start")).sum()
     ).select(
+        pl.col("chrom"),
         pl.col("genome"),
-        pl.col("genome_length")
+        (pl.col("A")+pl.col("C")+pl.col("G")+pl.col("T")).alias("coverage")
     )
-    profile=profile.with_columns(
-        pl.col("pos").shift(1).fill_null(0).over("chrom").alias("prev_pos"),
-    ).with_columns(
-        (pl.col("pos") - pl.col("prev_pos")).clip(lower_bound=1).alias("gap_size")
+    profile=profile.group_by("genome").agg(
+        total_covered_sites=pl.len(),
+        coverage=pl.col("coverage").sum()
     ).join(
-        stb,
-        left_on="chrom",
-        right_on="scaffold",
+        genome_lengths,
+        on="genome",
         how="left"
-    ).group_by("genome").agg(
-        cv=pl.col("gap_size").filter(pl.col("gap_size") > 1).std()/pl.col("gap_size").filter(pl.col("gap_size") > 1).mean(),
-        total_coverage=pl.col("coverage").sum(),
-        covered_positions=(pl.col("coverage")>0).sum()
     ).join(
-        genome_lengths,
+        genome_gap_stats,
         on="genome",
         how="left"
     ).with_columns(
-        (pl.col("covered_positions")/pl.col("genome_length")).alias("breadth"),
-        (pl.col("total_coverage")/pl.col("genome_length")).alias("coverage"),
-    ).select(
-        pl.col("genome"),
-        pl.col("cv"),
-        pl.col("breadth"),
-        pl.col("coverage"),
+        coverage=(pl.col("coverage")/pl.col("genome_length")),
+        breadth=(pl.col("total_covered_sites")/pl.col("genome_length")),
     ).with_columns(
-        (pl.col("breadth")/(1-(-0.883*pl.col("coverage")).exp())).alias("ber"),
+        ber=pl.col("breadth")/(1-(-0.883*pl.col("coverage")).exp()),
+        fug=pl.col("fug")
     ).with_columns(
-        pl.when(
-            pl.col("coverage") >= min_cov_constant_poisson
-        ).then(
-            pl.col("ber") >= ber
+    
+    pl.when(pl.col("coverage") > min_cov_use_fug)
+    .then(
+        pl.col("ber") > ber
         ).otherwise(
-            (pl.col("cv") <= cv_threshold)  & (~pl.col("ber").is_nan())
-        ).alias("is_present")
+            (pl.col("fug")/0.632 < fug) &
+            (pl.col("ber") > ber)
+        ).fill_null(False).alias("is_present"))
+
+    return profile.select(
+        pl.col("genome"),
+        pl.col("coverage"),
+        pl.col("breadth"),
+        pl.col("ber"),
+        pl.col("fug"),
+        pl.col("is_present")
     )
 
-    return profile
-        
+