gff3文件去重

```{python}

#https://www.stechies.com/read-file-line-by-line-python/

## https://blog.csdn.net/weixin_41578567/article/details/103262925

## https://www.statology.org/pandas-remove-duplicates-keep-max/

import os

import sys

import pandas as pd

input_file="alltransc.genome.sorted.uniq.gene+_gene-.bingji_145.sart1.haveid.2.genome.mRNA_CDS_1_head100.gff3"

## read big data

#df=pd.read_csv(input_file,header=None,sep="\t",nrows=100)

data=pd.read_csv(input_file,header=None,sep="\t", chunksize=500)

df = pd.concat(data)

## need install datatable module in conda environment by "conda install -c conda-forge datatable"

#import datatable as dt

#df = dt.fread()

num_list=list(range(1,10))

df.columns=["V"+str(x) for x in num_list]

#drop rows with duplicate team and positions but keeps row with max points

## V5=mRNA; V3=start; V4=end

df_new = df.sort_values('V5', ascending=False).drop_duplicates(['V3', 'V4']).sort_index()

df_new.to_csv("output.gff3",sep="\t",index=False,header=False)

```