参考：

https://www.jianshu.com/p/16dbbb824e3c

https://www.jianshu.com/p/fbcf00a46574

####安装

cd /home/yifan/software

wget --no-check-certificate https://sourceforge.net/projects/subread/files/subread-2.0.2/subread-2.0.2-Linux-x86_64.tar.gz

tar -xzvf subread-2.0.2-Linux-x86_64.tar.gz

cd subread-2.0.2-Linux-x86_64/bin/

ls -h

运行

/public/vip/biosoft/subread-2.0.2-Linux-x86_64/bin/featureCounts

####输入文件

1）reads的比对情况，这种信息通常都用BAM/ SAM文件来存储

2）区间注释文件，支持两种格式

在定量的时候，支持对单个feature 定量（对外显子定量），也支持对meta-feature 进行定量（对基因进行定量）， meta-feature 的定量是属于同一meta-features 下的所有features 的总和；

当reads 比对到2个或者以上的features 时，默认情况下，featureCounts在统计时会忽略到这部分reads, 如果你想要统计上这部分reads, 可以添加-O 参数，此时一条reads 比对到多个feature， 每个feature 定量时，都会加1，对于meta-features 来说，如果比对到多个features 属于同一个 meta-features(比如一条reads比对到了exon, 但这些exon 属于同一个gene), 则对于这个gene 而言，只会计数1次；

总之，不管对于feature 还是meta-feature, 只有 比对多个不同的区间时，才会分别计数；

######使用

##单个样本定量

featureCounts -T 5 -t exon -g gene_id -a annotation.gtf -o counts.txt mapping.sam

###多个样本归一化

featureCounts -t exon -g gene_id -a annotation.gtf -o counts.txt library1.bam library2.bam library3.bam

# 运行featureCounts对所有样本进行基因水平定量

gtf=/teach/database/gtf/gencode.v25.annotation.gtf.gz

featureCounts -T 1 -p \

-t exon -g gene_id \

-a $gtf -o all.id.txt *.sort.bam

#解释：-T 1：线程数为1；-p：表示数据为paired-end，双末端测序数据；-t exon表示 feature名称为exon；-g gene_id表示meta-feature名称为gene_id(ensembl名称)；-a $gtf 表示输入的GTF基因组注释文件；-o all.id.txt 设置输出文件类型；*.sort.bam是被分析的bam文件。featureCounts支持通配符*

# 只保留all.id.txt文件的【基因名】和【样本counts】

下面对几个常用的选项详细解释一下：

-F # 指定区间注释文件的格式，默认是GTF

-o # 输出文件:可输出raw counts的txt文本及raw counts的summary文本

-p # 针对paired-end数据

-g # 指定统计的meta-feature名称，默认是gene_id

-t # 指定统计的feature名称， 默认是exon。可有CDS,exon,five_prime_utr,gene,Selenocysteine,start_codon, stop_codon,three_prime_utr,transcript

-T # 指定线程数

-a  :  指定的区间注释文件，默认是gtf格式

-T  :  线程数，默认是1

-t  :  想要统计的feature 的名称， 取值范围是gtf 文件中的第3列的值，默认是exon

-g  :  想要统计的meta-feature 的名称，取值范围参考gtf 第9列注释信息，gtf 的第9列为 key=value 的格式， -g 参数可能的取值就是所有的key, 默认值是gene_id。准备gtf 文件时，要确保-g 参数指定的值都是唯一标识符，才能达到预期的效果；

####输出结果解读

Geneid：基因的ensemble基因号；

Chr：多个feature所在的染色体编号;

Start：多个feature起始位点,与前面一一对应

End：多个feature终止位点,与前面一一对应

Strand：正负链

Length：基因长度

sampleID：一列代表一个样本，数值表示比对到该基因上的read数目

###MultiQC分析featureCounts结果

multiqc all.id.txt.summary #对featureCounts的结果进行整合，html文件可视化

###########own

cd /home/yifan/project/CZM/exo.m6A/Data/CleanData/fastp.results/4.bowtie2.results

gtf=/home/yifan/data/ref/mouse/Mus_musculus.GRCm38.90.gtf

for i in LR22A21FY{19..24}

do

featureCounts -T 20 -p \

-t exon -g gene_id \

-a $gtf -o input.count.txt ${i}.sorted.bam

done

# 只保留all.id.txt文件的【基因名】和【样本counts】

cat all.id.txt | cut -f1,7- > counts.txt

###########帮助文档

Usage: featureCounts [options] -a <annotation_file> -o <output_file> input_file1 [input_file2] ...

## Mandatory arguments:

  -a <string>        Name of an annotation file. GTF/GFF format by default. See

                      -F option for more format information. Inbuilt annotations

                      (SAF format) is available in 'annotation' directory of the

                      package. Gzipped file is also accepted.

  -o <string>        Name of output file including read counts. A separate file

                      including summary statistics of counting results is also

                      included in the output ('<string>.summary'). Both files

                      are in tab delimited format.

  input_file1 [input_file2] ...  A list of SAM or BAM format files. They can be

                      either name or location sorted. If no files provided,

                      <stdin> input is expected. Location-sorted paired-end reads

                      are automatically sorted by read names.

## Optional arguments:

# Annotation

  -F <string>        Specify format of the provided annotation file. Acceptable

                      formats include 'GTF' (or compatible GFF format) and

                      'SAF'. 'GTF' by default.  For SAF format, please refer to

                      Users Guide.

  -t <string>        Specify feature type(s) in a GTF annotation. If multiple

                      types are provided, they should be separated by ',' with

                      no space in between. 'exon' by default. Rows in the

                      annotation with a matched feature will be extracted and

                      used for read mapping.

  -g <string>        Specify attribute type in GTF annotation. 'gene_id' by

                      default. Meta-features used for read counting will be

                      extracted from annotation using the provided value.

  --extraAttributes  Extract extra attribute types from the provided GTF

                      annotation and include them in the counting output. These

                      attribute types will not be used to group features. If

                      more than one attribute type is provided they should be

                      separated by comma.

  -A <string>        Provide a chromosome name alias file to match chr names in

                      annotation with those in the reads. This should be a two-

                      column comma-delimited text file. Its first column should

                      include chr names in the annotation and its second column

                      should include chr names in the reads. Chr names are case

                      sensitive. No column header should be included in the

                      file.

# Level of summarization

  -f                  Perform read counting at feature level (eg. counting

                      reads for exons rather than genes).

# Overlap between reads and features

  -O                  Assign reads to all their overlapping meta-features (or

                      features if -f is specified).

  --minOverlap <int>  Minimum number of overlapping bases in a read that is

                      required for read assignment. 1 by default. Number of

                      overlapping bases is counted from both reads if paired

                      end. If a negative value is provided, then a gap of up

                      to specified size will be allowed between read and the

                      feature that the read is assigned to.

  --fracOverlap <float> Minimum fraction of overlapping bases in a read that is

                      required for read assignment. Value should be within range

                      [0,1]. 0 by default. Number of overlapping bases is

                      counted from both reads if paired end. Both this option

                      and '--minOverlap' option need to be satisfied for read

                      assignment.

  --fracOverlapFeature <float> Minimum fraction of overlapping bases in a

                      feature that is required for read assignment. Value

                      should be within range [0,1]. 0 by default.

  --largestOverlap    Assign reads to a meta-feature/feature that has the

                      largest number of overlapping bases.

  --nonOverlap <int>  Maximum number of non-overlapping bases in a read (or a

                      read pair) that is allowed when being assigned to a

                      feature. No limit is set by default.

  --nonOverlapFeature <int> Maximum number of non-overlapping bases in a feature

                      that is allowed in read assignment. No limit is set by

                      default.

  --readExtension5 <int> Reads are extended upstream by <int> bases from their

                      5' end.

  --readExtension3 <int> Reads are extended upstream by <int> bases from their

                      3' end.

  --read2pos <5:3>    Reduce reads to their 5' most base or 3' most base. Read

                      counting is then performed based on the single base the

                      read is reduced to.

# Multi-mapping reads

  -M                  Multi-mapping reads will also be counted. For a multi-

                      mapping read, all its reported alignments will be

                      counted. The 'NH' tag in BAM/SAM input is used to detect

                      multi-mapping reads.

# Fractional counting

  --fraction          Assign fractional counts to features. This option must

                      be used together with '-M' or '-O' or both. When '-M' is

                      specified, each reported alignment from a multi-mapping

                      read (identified via 'NH' tag) will carry a fractional

                      count of 1/x, instead of 1 (one), where x is the total

                      number of alignments reported for the same read. When '-O'

                      is specified, each overlapping feature will receive a

                      fractional count of 1/y, where y is the total number of

                      features overlapping with the read. When both '-M' and

                      '-O' are specified, each alignment will carry a fractional

                      count of 1/(x*y).

# Read filtering

  -Q <int>            The minimum mapping quality score a read must satisfy in

                      order to be counted. For paired-end reads, at least one

                      end should satisfy this criteria. 0 by default.

  --splitOnly        Count split alignments only (ie. alignments with CIGAR

                      string containing 'N'). An example of split alignments is

                      exon-spanning reads in RNA-seq data.

  --nonSplitOnly      If specified, only non-split alignments (CIGAR strings do

                      not contain letter 'N') will be counted. All the other

                      alignments will be ignored.

  --primary          Count primary alignments only. Primary alignments are

                      identified using bit 0x100 in SAM/BAM FLAG field.

  --ignoreDup        Ignore duplicate reads in read counting. Duplicate reads

                      are identified using bit Ox400 in BAM/SAM FLAG field. The

                      whole read pair is ignored if one of the reads is a

                      duplicate read for paired end data.

# Strandness

  -s <int or string>  Perform strand-specific read counting. A single integer

                      value (applied to all input files) or a string of comma-

                      separated values (applied to each corresponding input

                      file) should be provided. Possible values include:

                      0 (unstranded), 1 (stranded) and 2 (reversely stranded).

                      Default value is 0 (ie. unstranded read counting carried

                      out for all input files).

# Exon-exon junctions

  -J                  Count number of reads supporting each exon-exon junction.

                      Junctions were identified from those exon-spanning reads

                      in the input (containing 'N' in CIGAR string). Counting

                      results are saved to a file named '<output_file>.jcounts'

  -G <string>        Provide the name of a FASTA-format file that contains the

                      reference sequences used in read mapping that produced the

                      provided SAM/BAM files. This optional argument can be used

                      with '-J' option to improve read counting for junctions.

# Parameters specific to paired end reads

  -p                  If specified, libraries are assumed to contain paired-end

                      reads. For any library that contains paired-end reads, the

                      'countReadPairs' parameter controls if read pairs or reads

                      should be counted.

  --countReadPairs    If specified, fragments (or templates) will be counted

                      instead of reads. This option is only applicable for

                      paired-end reads. For single-end data, it is ignored.

  -B                  Only count read pairs that have both ends aligned.

  -P                  Check validity of paired-end distance when counting read

                      pairs. Use -d and -D to set thresholds.

  -d <int>            Minimum fragment/template length, 50 by default.

  -D <int>            Maximum fragment/template length, 600 by default.

  -C                  Do not count read pairs that have their two ends mapping

                      to different chromosomes or mapping to same chromosome

                      but on different strands.

  --donotsort        Do not sort reads in BAM/SAM input. Note that reads from

                      the same pair are required to be located next to each

                      other in the input.

# Number of CPU threads

  -T <int>            Number of the threads. 1 by default.

# Read groups

  --byReadGroup      Assign reads by read group. "RG" tag is required to be

                      present in the input BAM/SAM files.

# Long reads

  -L                  Count long reads such as Nanopore and PacBio reads. Long

                      read counting can only run in one thread and only reads

                      (not read-pairs) can be counted. There is no limitation on

                      the number of 'M' operations allowed in a CIGAR string in

                      long read counting.

# Assignment results for each read

  -R <format>        Output detailed assignment results for each read or read-

                      pair. Results are saved to a file that is in one of the

                      following formats: CORE, SAM and BAM. See Users Guide for

                      more info about these formats.

  --Rpath <string>    Specify a directory to save the detailed assignment

                      results. If unspecified, the directory where counting

                      results are saved is used.

# Miscellaneous

  --tmpDir <string>  Directory under which intermediate files are saved (later

                      removed). By default, intermediate files will be saved to

                      the directory specified in '-o' argument.

  --maxMOp <int>      Maximum number of 'M' operations allowed in a CIGAR

                      string. 10 by default. Both 'X' and '=' are treated as 'M'

                      and adjacent 'M' operations are merged in the CIGAR

                      string.

  --verbose          Output verbose information for debugging, such as un-

                      matched chromosome/contig names.

  -v                  Output version of the program.