MOD=998244353
class Graph():
    def __init__(self,size,directed=False):
        self.dir=directed
        self.size=size
        self.gr=[[] for i in range(size)]
        self.edges=[]
    
    def add_edge(self,u,v,status={}):
        self.gr[u].append(self.Edge(u,v,status))
        if not(self.dir):
            self.gr[v].append(self.Edge(v,u,status))
        self.edges.append(self.Edge(u,v,status))

    def node(self,v):
        return self.gr[v]
    
    class Edge():
        def __init__(self,st,to,status):
            self.st=st
            self.to=to
            self.status=status
        
        def __getitem__(self,val):
            return self.status[val]

from collections import deque
def bridge(gr):
    pre=[-1]*gr.size
    low=[-1]*gr.size
    ret=[]
    for i in range(gr.size):
        if pre[i]!=-1:
            continue
        vert=deque([[0,i,-1,-1]])
        cnt=0
        while len(vert)>0:
            step,now,bef,ed=vert[-1]
            if step==1:
                for j in gr.node(now):
                    if j.to!=bef:
                        low[now]=min(low[now],low[j.to])
                if ed!=-1 and pre[now]<=low[now]:
                    ret.append(gr.node(bef)[ed])
                vert.pop()
                continue
            if pre[now]!=-1:
                low[bef]=min(low[bef],low[now])
                vert.pop()
                continue
            pre[now]=cnt
            low[now]=cnt
            cnt+=1
            vert[-1][0]=1
            for j,k in enumerate(gr.node(now)):
                if k.to==bef:
                    continue
                if low[k.to]==-1:
                    vert.append([0,k.to,now,j])
                else:
                    low[now]=min(low[now],low[k.to])
    return ret

N,K=map(int,input().split())
namori=Graph(N)
for i in range(N):
    u,v=map(int,input().split())
    namori.add_edge(u-1,v-1)
cycle=N-len(bridge(namori))
dp=[[0,0] for i in range(cycle)]
dp[0][0]=K
for i in range(cycle-1):
    dp[i+1][0]=dp[i][1]
    dp[i+1][1]=(dp[i][0]*(K-1)+dp[i][1]*(K-2))%MOD
print((dp[-1][1]*pow(K-1,N-cycle,MOD))%MOD)