• 利用一维卷积做Kmer提取，可实现多组Kmer叠加
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
class K_mer_aggregate(nn.Module):
    def __init__(self,kmers,in_dim,out_dim,dropout=0.1):
        super(K_mer_aggregate, self).__init__()
        self.dropout=nn.Dropout(dropout)
        self.convs=[]
        for i in kmers:
            print(i)
            # sequence_length -> sequence_length-i+1
            self.convs.append(nn.Conv1d(in_dim,out_dim,i,padding=0))
        self.convs=nn.ModuleList(self.convs)
        self.activation=nn.ReLU(inplace=True)
        self.norm=nn.LayerNorm(out_dim)

    def forward(self,x):
        # 卷积是在最后一个维度上做的
        # (batch_size, sequence_length, features)->(batch_size, features, sequence_length)
        x = x.permute(0,2,1)
        outputs=[]
        for conv in self.convs:
            outputs.append(conv(x))
        outputs=torch.cat(outputs,dim=2)
        outputs=self.norm(outputs.permute(0,2,1))
        return outputs
  • 实例
1
2
3
4
5
6
7
8
9
10
kmers = [3,6]
ninp = 128
K  = K_mer_aggregate(kmers,ninp,ninp)

x = torch.rand((5, 10, 128))*10
out = K(x)
out.shape
>>>
# 13 = (10-3+1) + (10-6+1)
torch.Size([5, 13, 128])